Aqarios Launches Constraint-Native Quantum Optimizer in the IBM Qiskit Functions Catalog
Aqarios released a constraint‑native quantum optimizer in IBM’s Qiskit Functions Catalog, showing benchmarked performance on IBM Heron processors.

Introduction
Aqarios GmbH, a Munich‑based quantum‑software spin‑off of LMU Munich, has added its Constrained Quantum Optimizer to the IBM Qiskit Functions Catalog. The tool targets users of IBM Quantum Premium, Flex, and on‑premises systems, offering a managed solver for discrete combinatorial optimization problems.
Native Constraint Handling
Traditional quantum optimization pipelines translate constraints into penalty terms, expanding the search space and requiring careful weight tuning. Aqarios’ approach embeds hard one‑hot and set‑packing constraints directly into the quantum circuit using its Iterative Warm‑Start XY‑Mixer QAOA (IWS‑QAOA). This keeps the state evolution confined to the feasible subspace, eliminating the need for variational parameter training loops on hardware.
Benchmark Results on IBM Heron
The optimizer was evaluated on several IBM Heron processors (e.g., ibm_boston and ibm_marrakesh). Highlights include:
- Hardware‑Tailored Spin Glass (144‑qubit): exact optima in 3 of 5 instances, >99 % approximation on the rest.
- Maximum Independent Set (MIS) problems up to 156 qubits: 100 % feasible optimum with runtimes around 100–200 s total (≈30 s QPU time after warm‑start optimization).
- Karate Club Max‑4‑Cut (156‑qubit): 100 % feasible optimum in 200 s total runtime.
These results show a reduction of QPU compute time from 130–200 s to roughly 30 s for medium and large problem tiers.
Practical Workflow
Before circuit synthesis, an automated pre‑processor performs graph‑clique merging and pendant‑node fixing to compress qubit usage. After execution, a greedy steepest‑descent post‑processor repairs any noise‑induced constraint violations. The end‑to‑end pipeline is available as a ready‑to‑use function in the Qiskit catalog, simplifying adoption for quantum‑optimization practitioners.
Closing Thoughts
Aqarios’ optimizer demonstrates that constraint‑native quantum algorithms can deliver tangible speed‑ups on current superconducting hardware. By removing penalty‑based overhead and variational training, the tool makes quantum‑enhanced combinatorial optimization more accessible to IBM’s user base, marking a practical step forward for near‑term quantum applications.
Hype check
RealThe announcement is backed by a peer‑reviewed New Journal of Physics paper and includes concrete benchmark data on IBM hardware, matching the headline claim.
- Peer-reviewed paper, or press release only?Paper — Validated in New Journal of Physics
- Error-corrected (logical) qubits, or physical qubits only?Logical / not a qubit claim — Not a qubit‑count claim
- A new milestone, or a rebrand / restatement?New — New optimizer with benchmark results
- Independent confirmation?Not yet — No external verification beyond the authors
- Does the headline claim match what the source shows?Matches — Headline accurately reflects the optimizer launch
An editorial assessment of the claim as reported, not a judgment of the company or researchers. How we rate hype →
Source
Read original sourceWhy we picked this
Accepted from curated QuantumUnzipped source (relevance scoring disabled).