Insider Brief
- Qubit Pharmaceuticals and Sorbonne University demonstrated that quantum computers can outperform classical machines for non-reversible Markov chains, overturning a long-standing theoretical speed limit on quantum advantage.
- The study shows that quantum algorithms can achieve greater-than-quadratic, and in some cases up-to-exponential, speedups for one-way processes such as chemical reactions, protein folding, heat flow, and financial dynamics.
- The results extend quantum acceleration from idealized reversible systems to real-world irreversible systems, with potential impacts on drug discovery, materials science, and risk modeling.
PRESS RELEASE — Qubit Pharmaceuticals, in collaboration with Sorbonne University, has demonstrated that quantum computers can perform certain calculations faster than scientists believed physically possible. The discovery overturns one of the field’s longest-standing theoretical limits and redefines what quantum advantage means.
Their study, published in Nature Communications, shows that quantum machines can surpass classical computers when dealing with non-reversible Markov chains. These are systems that describe processes moving in one direction, such as chemical reactions, protein folding, or heat flow. The development extends the previously known quadratic speedup to cases where even greater, up-to-exponential gains are possible, unlocking new potential in
fields from drug discovery and materials science to finance and risk modeling.
“Quantum computers were expected to provide a quadratic speedup in the reversible setting,” said Baptiste Claudon, lead author and part of Qubit Pharmaceuticals’ research team. “By studying real-world systems, we found that the quadratic ceiling could be broken.”
Beyond the classical speed limit
In practice, a Markov chain is a sequence of possible events where the probability of the next step depends only on the current state, which is a cornerstone of modern statistical physics and machine learning. Until now, quantum algorithms could only handle reversible versions of these chains: clean, symmetric systems that can be mathematically “rewound.”
But the real world doesn’t run backward. Molecules react, markets shift, and heat dissipates — processes that are non-reversible by nature. Modeling them accurately has been one of the hardest challenges in computational physics.
This study shows, for the first time, that quantum algorithms can handle both reversible and non-reversible Markov chains, breaking through the long-assumed symmetry barrier.
“This bridges a critical gap between quantum computing and the physical world,” said Professor Jean-Philip Piquemal, co-author and co-founder of Qubit Pharmaceuticals. “Most natural and economic phenomena are irreversible — and now quantum algorithms can finally model them directly, with a clear computational advantage.”
A new quantum framework for irreversible systems
The paper introduces two new techniques that allow quantum computers to follow the natural direction of complex systems without having to “rewind” time. In practical terms, this means quantum computers can simulate chemical reactions or financial markets as they actually unfold, not as idealized two-way systems.Building on this, the authors lay out the first full framework for accelerating such one-way systems using quantum operations already proven in today’s most advanced algorithms — a clear blueprint for applying the discovery to real-world simulations and data problems.
In measurable terms, the approach could allow quantum computers to perform billions of classical iterations in just thousands, dramatically reducing the time required for large-scale simulations in drug design, materials discovery, and quantum-AI-based optimization.
“This work lets quantum computers speak the same language as the physical world,” said Robert Marino, CEO and founder of Qubit Pharmaceuticals. “Most real processes only move forward — and now quantum algorithms can follow that flow instead of fighting against it.”
From theory to real-world impact
Where earlier quantum algorithms offered a quadratic advantage in reversible systems, this framework generalizes that advantage to the non-reversible realm — suggesting that up-to-exponential speedups may occur in certain cases. These results are expected to influence how statistical physics methods are implemented across disciplines, while opening the door to a new class of quantum applications.
For Qubit Pharmaceuticals, the discovery also strengthens the company’s proprietary simulation platform, which integrates quantum algorithms into practical molecular and materials modeling workflows. By bridging advanced research and applied computation, Qubit is positioning itself among Europe’s leading companies translating theoretical breakthroughs into commercial quantum advantage.
The work highlights the strength of Europe’s quantum ecosystem, combining expertise from Sorbonne Université, CNRS, and Qubit Pharmaceuticals, and supported by the European Research Council (ERC) and France’s PEPR EPIQ – Quantum Software and HQI programs — initiatives that position Europe at the forefront of quantum software and algorithm development.
“This result shows how academic research and private deep-tech efforts can jointly advance the frontiers of quantum science,” added Professor Piquemal. “It’s a tangible step toward real quantum advantage in chemistry, biology, and finance.”
About the Publication
The paper “Quantum Speedup for Nonreversible Markov Chains” Communications in October 2025.
LINK: https://doi.org/10.1038/s41467-025-65761-5 was published in Nature
