Quantum Benchmark Zoo

← All benchmarks

BACQ

Application-level

MetriQs-France benchmark suite scoring whole quantum stacks on optimization, linear systems, many-body simulation, and factoring, aggregated into one user-weighted figure of merit.

Measures
Sub-scores on four application pillars plus energy efficiency, aggregated into a user-preference-weighted global figure of merit
Introduced by
BACQ consortium coordinated by Thales (with Eviden, CEA, CNRS, Teratec & LNE) (2023)
Also known as
Application-oriented Benchmarks for Quantum Computing, Benchmarks for Application-Centric Quantum Computing

BACQ (Application-oriented Benchmarks for Quantum Computing) is a French benchmark suite that scores a whole quantum computing stack on reference application problems and aggregates the results into a single, user-weighted figure of merit. It is the flagship benchmarking effort of the MetriQs-France program under France 2030: a consortium coordinated by Thales with Eviden, CEA, CNRS, Teratec, and the national metrology lab LNE, announced in May 2023 with roughly 4 million EUR over three years and publicly defined in a March 2024 paper.

How it works

BACQ scores four application pillars, plus energy efficiency as a resource cost: combinatorial optimization (a MaxCut benchmark that adopts and extends Eviden’s Q-score), linear-system solving, quantum many-body physics simulation, and prime factorization. The sub-scores are then aggregated into a global figure of merit with MYRIAD, Thales’ multi-criteria decision-analysis tool, which weights the pillars by user preferences instead of forcing one universal ranking; Labreuche’s QUEST-IS 2025 paper works out scale-invariance properties of that aggregation.

Strengths and limitations

The multi-criteria aggregation is the distinctive contribution: no single number serves everyone, so different users can weight optimization against simulation or factoring. Suites like the QED-C benchmarks, by contrast, publish per-application results without an aggregate. The trade-off is maturity: as of July 2026 BACQ has no dedicated public code repository (the open-source code the project cites is Q-score’s) and no published machine ranking; the consolidated suite, leaderboard, and final results are still rolling out. The acronym’s expansion also varies by source (Eviden materials used “Benchmarks for Application-Centric Quantum Computing”), and BACQ should not be confused with Q-score itself, which it subsumes as one pillar.

Notable results

The funded project runs September 2023 through 2026, with international TQCI seminars in June 2024 and June 2025, and BACQ papers presented at QUEST-IS 2025 appeared in the Springer proceedings in 2026. The January 2026 Many-body Quantum Score (from Eviden/CEA authors overlapping BACQ’s team, matching its many-body pillar) stops short of billing itself as a BACQ deliverable.

Key papers

  • Q-score : Atos/Eviden's application-level metric: the largest MaxCut instance a quantum system can solve effectively, judged by beating random guessing by a set fraction of optimal-solver scaling.
  • QED-C Application-Oriented Benchmarks : Open-source consortium suite that scores devices on ~16 algorithm workloads, from Bernstein-Vazirani to VQE and MaxCut, via normalized fidelity on width-by-depth volumetric plots.