A new quantum algorithm successfully performed a 15-step nonlinear fluid simulation around a solid obstacle using real quantum hardware, marking the most physically complex publicly documented demonstration of its type. This method lowers qubit needs and circuit depth, edging industrial CFD applications closer to reality.
Finnish simulation company Quanscient and quantum middleware developer Haiqu have achieved what they claim to be the most physically complex quantum computational fluid dynamics simulation to date on actual hardware.
The companies conducted a 15-step nonlinear fluid simulation around a solid obstacle, such as fluid flow around a shape relevant to designing aircraft wings or vehicle aerodynamics, on IBM’s Heron R3 quantum computer using a newly developed algorithm known as the One-Step Simplified Lattice Boltzmann Method (OSSLBM).
Computational fluid dynamics (CFD) is one of the most resource-heavy areas in engineering simulation, requiring massive classical computing power to model fluid behavior around intricate shapes, with exponentially increasing demands as simulations increase in detail.
Quantum computing has been considered a promising route to overcome classical limits in simulations, but harnessing this potential has been challenging due to the extensive number of qubits and circuit depth required to run even moderately complex scenarios without error-prone calculations.
The OSSLBM algorithm tackles this problem by building on the quantum Lattice Boltzmann Method (QLBM), a recognized approach for translating classical fluid equations into quantum computations. This new framework lessens the computational load for each step, enabling longer multi-step simulations on current quantum hardware.
Haiqu’s middleware was crucial, reducing circuit depth, creating new algorithmic subroutines, and employing targeted error-reduction strategies to enable a workflow that would otherwise be unattainable with today’s devices.
The breakthrough lies in addressing fluid behavior interactions with solid obstacles, unlike past quantum CFD demonstrations that have mainly addressed simpler linear scenarios.
Modeling fluid movement around objects is essential for any significant industrial application. Professor Oleksandr Kyriienko remarked on its significance as an “interesting and timely contribution to quantum CFD,” highlighting the necessity for more research to achieve industrially relevant quantum solutions.
Quanscient and Haiqu have been working on quantum CFD since at least 2024, being finalists in the Airbus and BMW Quantum Mobility Challenge, with prior projects on IonQ hardware via Amazon Braket. While industrial applications are still years away, the current work serves as a research milestone proving that this approach is viable on existing hardware at this complexity level.