← Latest papers
🌀 nonlinear sciences

Nonlinearity-preserving quantum simulation of flows in the early fault-tolerant regime

This paper proposes a nonlinearity-preserving quantum simulation framework that couples sparse-identified polynomial reduced-order models with Walsh-synthesized Koopman-von Neumann transport to demonstrate the feasibility of simulating early fault-tolerant fluid flows, such as a cylinder wake, within hours using approximately 560,000 physical qubits while mitigating noise through reference circuits.

Original author: Hansol Noh (David), Unam Oh (David), Sangkyu Baek (David), Doyeol (David), Ahn

Published 2026-11-18
📖 5 min read🧠 Deep dive

Original author: Hansol Noh (David), Unam Oh (David), Sangkyu Baek (David), Doyeol (David), Ahn

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

Fluids are rarely still. From the swirling wake behind a ship's hull to the air rushing over an airplane wing, moving liquids and gases are defined by their chaos. When a fluid flows past an obstacle, such as a cylinder, it does not simply slide around it. Instead, the flow becomes unstable, breaking into a repeating pattern of spinning vortices that shed from the object. While a single path of fluid might seem predictable, the reality is that tiny, unavoidable uncertainties in the starting conditions can cause individual paths to diverge wildly. To understand such systems, scientists do not track just one line of flow; they track a whole cloud of possibilities, a probability distribution that shows where the fluid is likely to be. This cloud of possibilities evolves over time according to a specific set of rules. The challenge for modern computing is that simulating this entire cloud, rather than just a single path, is incredibly difficult. Classical computers struggle with the sheer number of calculations required to keep the cloud accurate as it stretches and squeezes.

Quantum computers offer a different way to handle this problem. Unlike classical machines that process information in bits of zero or one, quantum machines use qubits, which can exist in complex combinations of states. This allows them to represent probability clouds directly. However, building a quantum computer that is powerful enough to solve real-world fluid problems is a work in progress. We are currently in an "early fault-tolerant" era, where machines are large enough to run complex algorithms but still prone to errors that require constant correction. The question researchers face is whether these early machines can actually perform useful simulations of nonlinear fluid dynamics before the technology matures into a fully error-free supercomputer.

A team of researchers has now demonstrated a method to simulate the chaotic wake behind a cylinder using a quantum approach designed specifically for these early machines. They focused on a classic test case: the flow of fluid past a cylinder at a speed where the wake becomes unstable and settles into a rhythmic pattern of vortex shedding. Instead of trying to simulate every single point in the fluid, which would require too many resources, they first simplified the problem. They used a technique called sparse identification to find a compact mathematical description of the flow's essential behavior. This description, known as a reduced-order model, captures the growth and saturation of the wake using just a few key variables and simple polynomial relationships. By reducing the complexity of the physics first, they could translate the problem into a format that a quantum computer could handle without getting overwhelmed.

The researchers then mapped this simplified model onto a quantum circuit using a method called Koopman–von Neumann embedding. This approach treats the probability of the fluid's state as a wave, allowing the quantum computer to evolve the entire probability cloud at once. A key innovation in their work was a technique called Walsh synthesis, which streamlined the quantum operations needed to move the probability cloud forward in time. This method ensured that the number of quantum steps required did not explode as the simulation became more detailed. Instead, the cost grew very slowly, making the simulation feasible on machines with limited resources. The team estimated that to simulate this fluid wake with high accuracy, they would need a quantum computer with roughly 560,000 physical qubits. While this sounds like a massive number, it is within the range of what experts predict will be available in the near future, specifically by the late 2020s.

To prove their concept was not just theoretical, the team ran a smaller version of their simulation on actual quantum hardware provided by IBM. They used a four-qubit processor to simulate the radial movement of the fluid wake. Because the machine was noisy and prone to errors, they employed a clever trick: they ran reference circuits with the same structure but with the motion turned off. By comparing the noisy results of the real simulation against these reference runs, they could mathematically strip away the noise and reveal the true underlying signal. The experiment successfully showed that the quantum computer could track the fluid's probability cloud as it moved toward a stable, repeating pattern, confirming that the nonlinear dynamics were preserved despite the hardware's imperfections.

The study provides a clear roadmap for how to bridge the gap between abstract quantum algorithms and practical fluid simulations. The researchers found that by balancing the accuracy of the simplified model against the cost of the quantum circuit, they could design simulations that fit within the capabilities of early fault-tolerant machines. Their estimates suggest that a full, high-fidelity simulation of the cylinder wake could be completed in about 1.6 hours on a machine with 560,000 qubits. This is a significant milestone because it moves the field from asking "can we do this?" to "how do we build the machine to do this?" The work demonstrates that with the right combination of mathematical simplification and efficient quantum circuit design, we can begin to harness quantum computers to solve complex, nonlinear problems in fluid dynamics long before the technology reaches its full potential.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →