Quantum science, anticipating evolutions in the computer science of the future

Quantum technology is shaking up computer science and its applications, from cyber-security to digital simulation, optimisation and artificial intelligence, which are all strategic sectors for states and businesses. In Auvergne Rhône-Alpes, the second most prominent region after Ile-de-France in quantum sciences, Inria is acting with its academic and industrial partners to anticipate future evolutions in the digital sector related to quantum algorithmics and information theory.

The institute is thus to create a dedicated team in 2022. Based in both Grenoble and Lyon, the team will draw on the existing expertise of Inria, which has a long-standing interest in this field, and on that of its partners Grenoble-Alpes University, Claude Bernard – Lyon 1 University  and the École Normale Supérieure de Lyon (ENS). In particular, this team will count on the expertise of Omar Fawzi (as team leader) and Alastair Abbott, senior researcher and research scientist at Inria respectively.

Both researchers have been working for more than a decade on the theory of quantum information and are interested in the development of algorithms adapted to quantum computers. What are their backgrounds? 

How does quantum computing work ? This technology uses the physical properties of matter in infinitely small dimensions ; at these scales, quantum systems possess the unique ability to form a superposition of states. “This is the feature that a quantum computer uses ; the possibility to create superpositions of states which then allows algorithms to run faster, and sometimes exponentially so”, Omar Fawzi explains. Such is the case, for example, for the integer factorisation problem, which comes into play in certain methods of cryptography.”

The development of operational quantum computers raises numerous technical and scientific difficulties. For example, qubits, (the quantum analogue of the bit in classical computing) can lose their quantum properties due to interactions with their environment, which results in computational errors. This is referred to by scientists as “decoherence” in this case, and more generally as “noise”.

At present, computer prototypes contain a limited number of around fifty qubits. In theory, however, a much larger number would be required to correct these errors, which is impossible in practical terms for the moment. Based on mathematical techniques (known as “convex optimization”), Omar Fawzi and Alastair Abbott’s research is aimed at controlling and reducing the noise in quantum computers by encoding qubits in the most optimal way possible.

“Our work involves quantum noise in a broader context, that of quantum information theory”, Omar Fawzi points out.“In the last century, with the emergence and development of computer science and communication theory, Claude Shannon founded ‘information theory’, and with quantum information one must invent the appropriate concepts.”

Quantum computing thus requires a complete rethinking of computer science; machines, algorithms and software, etc., must be designed together in order to get the most out of this emerging technology. “We offer a global approach to the problem, by working with physicists, for example, in order to make use of the most accurate noise models”, Alastair Abbott tells us. “Having more realistic noise models, which are closer to physics, enables a fine-tuning of algorithmic choices, in other words, the way in which quantum operations will be programmed, equivalent to the ‘AND’, ‘OR’ functions, etc. in classical computing. A new way of programming could in turn influence the architecture choices of the physicists developing the machines.”

With the recruitment of two young research scientists, coming from Copenhagen and Zurich, the prospect of collaboration with Benjamin Huard, a researcher at the ENS Lyon Physics Laboratory and the numerous contacts forged during their past international experiences, Omar Fawzi and Alastair Abbott’s future team is taking a decidedly collaborative approach. “Some key players in quantum science are located around Lyon and Grenoble, and are already structured around major academic fields (physics, mathematics, human sciences and philosophy) and businesses. We provide the algorithmic aspect”, says Alastair Abbott.

The creation of this new Inria team forms part of a wider objective, the Quantum Plan, an ambitious national strategy for quantum computing,  which aims to rally, around the three major centres of CNRS, CEA and Inria, other national players in the field (universities, industries, research institutes, small and medium-sized enterprises and start-ups, etc.).

If one thing is certain, it is that states with a mastery of quantum computing will take a decisive lead in strategic fields (such as energy, defence and health, etc.). In addition, there is no need to wait to have quantum computers to develop the algorithms and study how to program them. Inria is particularly well positioned to move forward in this algorithmic field.

• Où en sont les ordinateurs quantiques ? [Where are we at with quantum computers?], Arte, 25 September 2021.
• Débruiter le quantique, l’ERC Starting Grant d’Omar Fawzi , [Denoising quantum - the ERC Starting Grant for Omar Fawzi], CNRS, 11 December 2019
• Les enjeux de l’ordinateur quantique sur lequel la recherche se mobilise [The quantum computer issues research is acting to solve], France Bleu,15 March 2021.