A new study led by researchers at University College London has shown that combining quantum computing and artificial intelligence can significantly improve predictions of complex physical systems over long periods of time. Hybrid approaches perform better than leading models that rely solely on traditional computers.
The result is scientific progresscould enhance the simulation of the behavior of liquids and gases, known as fluid dynamics. These types of models are essential in areas such as climate science, transportation, medicine, and energy production.
Why quantum computing makes a difference
The increased accuracy appears to come from the way quantum computers process information. Unlike classical computers, which use bits set to 1 or 0, quantum computers use qubits, which can exist as 1, 0, or anywhere in between. Additionally, each qubit can influence other qubits, allowing a relatively small number of qubits to represent a huge number of possible states.
Professor Peter Coveney, senior author from the UCL Chemistry and Advanced Research Computing Centre, explained the challenge: “To make predictions about complex systems, we either run full simulations that can take weeks (often too long to be useful), or we use AI models that are faster but less reliable on longer timescales.
“Our AI model, based on quantum information, means we can provide more accurate predictions faster. Making predictions about fluid flow and turbulence is a fundamental scientific challenge, but it also has many application areas. Our method can be used for climate prediction, modeling blood flow and molecular interactions, or better designing wind farms to produce more energy.”
How hybrid quantum AI methods work
Quantum computers are widely expected to surpass the capabilities of classical machines, but so far their use in the real world has been limited. This new approach integrates quantum computing into specific stages of the AI training process.
AI models typically learn from large datasets generated through simulation or observation. In this case, the data is first processed by a quantum computer to identify significant statistical patterns that are stable over time. These patterns, known as invariant statistical properties, are used to guide the training of AI models that run on traditional supercomputers.
More accuracy with less memory
AI systems that leverage quantum information achieved approximately 20% higher accuracy compared to standard AI models that do not use quantum-derived patterns. It also maintained stable predictions over long periods of time, even when modeling chaotic systems.
Another big advantage was efficiency. This method requires hundreds of times less memory and is much more practical for large-scale simulations.
Quantum effects behind efficiency
This performance improvement comes from two distinctive features of quantum computing. Quantum entanglement allows qubits to influence each other regardless of distance, while superposition allows qubits to exist in multiple states at once until they are measured. The combination of these properties allows quantum systems to process vast amounts of information in a compact form.
Demonstrate practical quantum benefits
Lead author Maida Wang from UCL’s Center for Computational Sciences said: “Our new method appears to demonstrate a ‘quantum advantage’ in a practical way, meaning that quantum computers outperform what is possible with classical computing alone. These discoveries may stimulate the development of new classical approaches that achieve even higher accuracy, but our method… The next step will be to scale up this technique using larger datasets and apply it to real-world situations where problems such as the following are typically encountered.Furthermore, a provable theoretical framework will be proposed. ”
Co-lead author Xiao Xue, from UCL’s Advanced Research Computing, added: “This study demonstrates for the first time that quantum computing can be meaningfully integrated with classical machine learning methods to tackle complex dynamical systems, including fluid dynamics. It is exciting to see this type of ‘quantum-informed’ approach progressing towards practical application.”
Capturing the physics of complex systems
The researchers suggest that quantum computers are particularly suited to modeling these systems because of their compact representation of the underlying physics. Many complex systems exhibit behavior similar to quantum effects. This means that changes in one area can affect distant parts of the system, similar to entanglement.
Overcoming the limitations of current quantum hardware
Current quantum computers face challenges such as noise, errors, and interference, and often require large numbers of measurements. The new method avoids these problems by using a quantum computer only once during the workflow, rather than repeatedly exchanging data between quantum and classical systems.
Experiment contents and future possibilities
The study used a 20-qubit IQM quantum computer connected to powerful classical computing resources at the Leibniz Supercomputing Center in Germany.
To function, quantum computers must operate at extremely low temperatures, around -273 degrees Celsius (close to absolute zero and colder than any other temperature in the universe).
The research was funded by UCL and the UK Engineering and Physical Sciences Research Council (EPSRC), with additional support from the IQM Quantum Computer and the Leibniz Supercomputing Center in Munich.
As researchers continue to expand this approach, it could open the door to more accurate and efficient predictions across a wide range of scientific and engineering applications.
