Mixed Encodings meet budgets for over half of tested quantum objectives

Previously, quantum optimisation faced limitations in utilising available resources; exact “mixed” encodings now combine native interactions with Rosenberg auxiliaries to overcome these constraints. The approach identifies infinite families of objectives solvable using qubit counts, penalties or gates unavailable to either fully native or fully reduced methods. These “mixed” encodings strategically combine existing interactions with auxiliary components, new ways to translate complex problems for quantum computers.

The method overcomes limitations in qubit numbers or available operations that hinder standard methods when solving optimisation challenges. Instances exist where this mixed method is uniquely feasible, offering potential benefits beyond current capabilities while acknowledging further work is needed on solution quality itself. Translating complex problems for quantum computers has long been a goal; this “Hamiltonian encoding” converts a problem into the language of qubits, the basic units of quantum information.

The team developed new “mixed” encodings that strategically combine existing interactions with auxiliary components, extra ‘helper’ qubits added to simplify calculations, much like scaffolding supports construction. These mixed approaches organise and streamline interactions between qubits using what’s known as a direct-pair class, ensuring accuracy throughout processing similar to neatly packing items in pairs before boxing them up. This raises an important question: can these hybrid designs unlock solutions inaccessible to either fully native or fully simplified encoding schemes. Further technical details will explore how this is achieved and verified.

Rosenberg Auxiliaries Enable Feasible Solutions For Previously Inaccessible Quantum Objectives

A routing experiment showed sharp improvement in feasibility, retaining 52.9% of mixed-only budget points across five seeds, and exceeded performance compared to other approaches lacking established benchmarks; this indicates improved resource management during testing scenarios. Pointwise-exact “mixed” representations were created by deliberately replacing some interactions with Rosenberg auxiliaries while preserving native operations, unlocking infinite families of objectives solvable using resources unavailable to either fully native or reduced encodings. This allows solutions within qubit, penalty, and gate constraints previously considered impossible, though solution quality remains separate from achieving feasible designs.

These methods offer an alternative approach for translating complex problems onto quantum computers beyond standard encoding techniques. Retaining 52.9% of mixed-only budget points across five tests where direct comparisons were possible demonstrates improved computational resource management during problem-solving. The precise “mixed” representations are achieved through deliberate replacement of certain interactions with Rosenberg auxiliaries, additional variables simplifying calculations, while maintaining core native operations on quantum computers.

A shared penalty condition and closed form certify exactness at every stage of selective reduction; this allows infinite families to meet resource budgets that exclude both fully native and reduced endpoints within the direct-pair class. Compilation on 12-qubit line, ring, and grid targets found certified regions for 23 of 30 weighted maximum three-satisfiability instances, all 30 constructed instances, and 2 of 30 cubic spin-glass instances under stated conditions. Resource feasibility is therefore a sound reason to select mixed encoding, while solution quality requires separate evaluation.

Resourceful encodings extend problem reach despite current performance gaps

University of Southampton researchers demonstrate a pathway to solving problems previously out-of-reach due to hardware limitations; their work highlights an uncomfortable truth: simply *finding* a solution isn’t enough. Currently, these mixed encodings lag behind fully native approaches in achieving optimal objective values, expanding what’s possible within fixed resource budgets (qubits, penalties, gates) but not yet consistently delivering better answers than existing methods. Even with this gap, the research remains key work because it expands possibilities for complex quantum calculations.

The team from the University of Warwick and Southampton broadened capabilities by re-encoding complex calculations for quantum computers given limited qubits and computational steps. Replacing selected higher-order interactions with shared auxiliary components while retaining others extends beyond what is achievable using either fully native or entirely simplified encoding methods for certain optimisation problems. Establishing resource feasibility as a design criterion validates an alternative when qubit counts, penalties, or gate limits restrict conventional approaches; mixed representations can meet joint budgets, limits on resources like qubits, gates, depth, and penalty scale, that exclude both standard (native) and fully reduced endpoint encodings. Compilation tests on 12-qubit systems found certified regions where only a mixed representation was feasible across multiple instances of weighted maximum three-satisfiability and cubic spin-glass problems.

Researchers demonstrated a method to re-encode complex calculations for quantum computers by strategically combining native interactions with shared auxiliary components. This approach expands the range of solvable optimization problems within fixed resource constraints, specifically qubit counts, penalties, and gate limitations, achieving feasibility in cases where traditional or simplified encoding methods fail. While current results indicate fully native encodings currently achieve better objective values, expanding feasible problem sizes remains a valuable contribution to the field.

👉 More information
🗞 Exact Mixed-Locality Encodings for Resource-Constrained QAOA
✍️ Jingcong Liu, Yiding Su, Hafiz Muhammad Waseem, Bo Chen and Carsten Maple (University of Warwick); Ruojun Zhang (University of Southampton)
🧠 ArXiv: https://arxiv.org/abs/2611.15663

Stay current

See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.

Avatar of Muhammad Rohail T.

Latest Posts by Muhammad Rohail T.: