Quantum-classical hybrid: turbulence prediction accuracy +20%

Quantum-Classical Hybrid Boosts Turbulence Prediction Accuracy by 20%

Researchers from University College London (UCL) have developed a hybrid method combining quantum computing and artificial intelligence that more accurately predicts the behavior of chaotic systems, such as turbulent flows. The new approach requires 100 times less memory compared to classical models and demonstrates an accuracy improvement of about 20%. The work was published in a peer-reviewed journal and is based on experiments with a 20-qubit IQM quantum computer integrated with supercomputing resources.

How the Hybrid Method Works

The method is based on task division: the quantum computer is used only during the extraction of key statistical patterns from data, while the main model training is performed on a classical supercomputer. This approach avoids problems associated with noise and errors in modern quantum devices that arise during continuous operation.

The quantum computer analyzes time series and extracts invariant statistical properties—characteristics that persist over time even in chaotic systems. These properties are then passed to a classical machine learning model, which uses them as additional anchor points for prediction.

Why This Matters for Chaos Prediction

Chaotic systems, such as turbulence, are extremely sensitive to initial conditions. Errors in predictions grow rapidly, making long-term forecasts unreliable. Traditional methods require either expensive full simulations or simplified models that lose accuracy over long time intervals.

The new hybrid approach solves this problem by providing a more compact representation of data. Quantum bits (qubits), thanks to superposition and entanglement, can store more information in fewer states. This allows the model to capture long-range correlations in the system that are inaccessible to classical algorithms.

Key Advantages of the Method

• 20% improvement in accuracy compared to classical machine learning models.
• 100-fold reduction in memory requirements due to quantum data compression.
• Stable predictions over long time intervals.
• Practical feasibility on existing quantum hardware.

What Matters

• The hybrid approach uses the quantum computer only in one stage, avoiding noise issues.
• The method demonstrated a practical quantum advantage in a computational physics task.
• Results were obtained on a real 20-qubit quantum computer, not a simulator.
• Further research aims to test the method on more complex systems and develop a theoretical foundation.

Application Prospects

Although the authors do not make bold claims about an immediate revolution in climate modeling, the work paves the way for practical use of quantum computing in tasks requiring long-term forecasting of complex systems. This could find applications in:

• Hydrodynamics and aerodynamics (aircraft design, wind farms).
• Climate models (weather and climate change prediction).
• Medicine (blood flow modeling).
• Energy (combustion process optimization).

Industry Significance

The UCL work is one of the few examples where a quantum computer provides measurable benefit in a real computational task. This could stimulate the development of hybrid architectures where quantum devices perform specialized functions rather than attempting to replace classical computers entirely.

— Editorial Team