Atos announces Q-score, the only universal metrics to assess quantum performance and superiority

Paris, December 4, 2020

Atos introduces “Q-rating”, the initially common quantum metrics, applicable to all programmable quantum processors. Atos’ Q-rating measures a quantum system’s effectiveness at dealing with real-everyday living complications, all those which simply cannot be solved by common desktops, relatively than only measuring its theoretical functionality. Q-rating reaffirms Atos’ commitment to deliver early and concrete positive aspects of quantum computing. In excess of the earlier five several years, Atos has become a pioneer in quantum purposes as a result of its participation in industrial and tutorial partnerships and funded jobs, operating hand-in-hand with industrials to establish use-cases which will be in a position to be accelerated by quantum computing.

Faced with the emergence of a myriad of processor technologies and programming techniques, businesses seeking to commit in quantum computing need to have a dependable metrics to support them decide on the most productive route for them. Remaining hardware-agnostic, Q-rating is an objective, simple and honest metrics which they can rely on,” reported Elie Girard, Atos CEO. “Since the start of ‘Atos Quantum’ in 2016, the initially quantum computing industry plan in Europe, our purpose has remained the exact: advance the growth of industry and investigate purposes, and pave the way to quantum superiority.”

What does Q-rating evaluate?

Right now the number of qubits (quantum models) is the most popular figure of advantage for evaluating the functionality of a quantum method. On the other hand, qubits are volatile and vastly change in top quality (velocity, security, connectivity, and many others.) from one particular quantum technological know-how to one more (such as supraconducting, trapped ions, silicon and photonics), producing it an imperfect benchmark instrument. By concentrating on the skill to resolve nicely-recognised combinatorial optimization complications, Atos Q-rating will provide investigate centers, universities, enterprises and technological leaders with explicit, dependable, objective and similar results when solving real-entire world optimization complications.

Q-rating measures the precise functionality of quantum processors when solving an optimization trouble, agent of the close to-expression quantum computing period (NISQ – Noisy Intermediate Scale Quantum). To provide a body of reference for comparing functionality scores and retain uniformity, Q-rating relies on a typical combinatorial optimization trouble, the exact for all assessments (the Max-Minimize Difficulty, comparable to the nicely-recognised TSP – Travelling Salesman Difficulty, see under). The rating is calculated based mostly on the utmost number of variables in just such a trouble that a quantum technological know-how can enhance (ex: 23 variables = 23 Q-rating or Qs).

Atos will manage the publication of a annually listing of the most potent quantum processors in the entire world based mostly on Q-rating. Due in 2021, the initially report will involve precise self-assessments supplied by producers.

Dependent on an open access program package deal, Q-rating is created on 3 pillars:

  • Software driven: Q-rating is the only metrics method based mostly on close to-expression available quantum algorithms and measures a quantum system’s capacity to resolve realistic operational complications
  • Openness and relieve of use: Common and no cost, Q-rating positive aspects from Atos’ technological know-how-neutral tactic. Its program package deal, including instruments and methodology, does not require hefty computation electrical power to work out the metrics
  • Objectiveness and dependability: Atos combines a hardware-agnostic, technological know-how-agnostic tactic with a strong experience in algorithm layout and optimization obtained operating with major industry customers and technological know-how leaders in the quantum field​. The methodology applied to build Q-rating will be manufactured community and open to assessment.

A no cost program package, which permits Q-rating to be run on any processor will be available in Q1 2021. Atos invites all producers to run Q-rating on their technological know-how and publish their results.

Many thanks to the highly developed qubit simulation abilities of the Atos Quantum Studying Machine (Atos QLM), its potent quantum simulator, Atos is in a position to work out Q-rating estimates for numerous platforms. These estimates just take into account the traits publicly supplied by the producers. Final results variety close to a Q-rating of fifteen Qs, but development is fast, with an approximated average Q-rating courting from one particular yr back in the space of 10 Qs, and an approximated projected average Q-rating courting one particular yr from now to be previously mentioned 20 Qs.

Q-rating has been reviewed by the Atos Quantum Advisory Board, a team of worldwide experts, mathematicians and physicists authorities in their fields, which satisfied on
December 4, 2020.