Description du projet
L’informatique quantique dynamisée par une approche rationalisée de la correction d’erreurs quantiques
La cohérence quantique, ou superposition d’états mécaniques quantiques ne durant généralement qu’une fraction de seconde, permet la formation de deux ondes quantiques qui interfèrent de manière cohérente l’une avec l’autre. C’est la base de l’informatique quantique et les scientifiques ont donc développé des techniques de correction d’erreurs quantiques pour protéger les informations quantiques des erreurs dues à la décohérence. Les solutions actuelles sont très gourmandes en ressources et vont créer des obstacles à la mise à l’échelle pour les applications de nouvelle génération et à une utilisation généralisée. Le projet SuperProtected, financé par l’UE, va développer une solution qui pourrait augmenter le temps de cohérence d’un facteur 100 tout en simplifiant considérablement les exigences matérielles.
Objectif
The quantum computer dream is driven by promises of unprecedented capabilities but is also facing a stark reality: quantum coherence is as powerful as it is difficult to protect. Quantum Error Correction (QEC) aims to extend coherence using redundancies but leads to solutions that are extremely resource-intensive: at present, protecting one bit of information requires at least ten thousand physical qubits. The main objective of this proposal is to engineer a new type of superconducting qubit, which will be intrinsically protected against de-coherence. Instead of matching the qubit states to the number of Cooper pairs or flux quanta of a given circuit, as is usually done, SuperProtected will exploit a completely new encoding scheme: quantum information will be stored as the parity of the number of Cooper pairs. This will be achieved by building a circuit component where charge transport occurs as pairs of Cooper pairs (4e-tunneling) while the standard single pair transport (2e-tunneling), or Josephson current, is reduced to zero. This new paradigm implies inductances with unprecedented value (10μH), also known as superinductances. The novel approach builds on two technological steps: a new high-kinetic inductance superconductor (InOx) and suspended silicon membranes. Improvement of the coherence time over current state-of-the-art is expected to be two orders of magnitude. The proposed qubits offer another major advantage: protected gates can be implemented using a simple modification of the architecture. The resulting protected qubit will extend the frontiers of the current knowledge in QEC and bring down the hardware requirements for a logical qubit by several orders of magnitude. Such a result would considerably change the quantum computing landscape.
Champ scientifique
Programme(s)
Régime de financement
ERC-COG - Consolidator GrantInstitution d’accueil
75794 Paris
France