Quantinuum’s Correlation-Based Optimization Solves Problems Up To 1000 Variables

Quantinuum, along with Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, and Konstantinos Meichanetzidis, has demonstrated a new approach to optimization, solving problems with tens to thousands of variables using a correlation-based method. The work centers on Pauli Correlation Encoding, a framework where classical variables are encoded into many-body Pauli observables, but with a surprising twist: the necessary calculations can be performed classically. Researchers instantiated this using free-fermionic evolutions and Instantaneous Quantum Polynomial circuits, producing high-quality solutions across benchmarks including MaxCut, Maximum Independent Set, Multi-Dimensional Knapsack, and Max3SAT. As stated in the paper, this yields a dequantized baseline for evaluating future quantum implementations of PCE, revealing that PCE is naturally understood as a correlation-based optimization framework with both quantum and classically simulable realizations.

Solving problems ranging from tens to thousands of variables using a novel approach to encoding information, Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, and Quantinuum recently demonstrated a surprising capability within quantum-inspired optimization. This means the team has established a classical benchmark against which to measure the true benefit of employing quantum hardware for this type of optimization. On MaxCut, Maximum Independent Set, Multi-Dimensional Knapsack, and Max3SAT benchmarks, these methods produce high-quality solutions.

Demonstrating efficient classical simulation of certain implementations, Pauli Correlation Encoding (PCE) exploration by Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, and Quantinuum presents a surprising turn in the pursuit of quantum optimization. While many quantum approaches aim to surpass classical capabilities, this work establishes a dequantized baseline against which future quantum implementations of PCE can be rigorously evaluated. This baseline is achieved through realizations where all necessary expectation values are classically computable. Researchers instantiated this using both free-fermionic evolutions, realized by matchgate circuits, and Instantaneous Quantum Polynomial (IQP) circuits, effectively bypassing the need for quantum hardware to generate initial solutions. The team successfully tackled optimization problems ranging from tens to thousands of variables across established benchmarks, MaxCut, Maximum Independent Set, Multi-Dimensional Knapsack Problem, and Max3SAT, producing high-quality solutions.

Achieving competitive results with entirely classical simulations of a quantum-inspired framework is being explored by Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, and Quantinuum researchers as a surprising avenue in quantum optimization. Their work centers on Pauli Correlation Encoding (PCE), a method for tackling binary optimization problems by encoding classical variables into the expectation values of quantum observables. This dequantized approach leverages circuit families, specifically, matchgate circuits realizing free-fermionic evolutions and Instantaneous Quantum Polynomial (IQP) circuits, that allow for polynomial-time evaluation of the required correlations. On MaxCut, Maximum Independent Set, Multi-Dimensional Knapsack, and Max3SAT benchmarks, these methods produce high-quality solutions across problem sizes ranging from tens to thousands of variables.

Achieving competitive results entirely through classical computation, Quantinuum, along with Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, has introduced a surprising development in quantum-inspired optimization. While Pauli Correlation Encoding (PCE) typically demands quantum hardware to estimate crucial values, researchers have demonstrated realizations where these calculations can be performed using standard computers. This scale represents a significant advance, producing high-quality solutions across problem sizes ranging from tens to thousands of variables. The researchers emphasize that this isn’t about replacing quantum approaches, but rather establishing a clear performance standard, as stated in the paper: “This yields a dequantized baseline for evaluating future quantum PCE implementations.”

The pursuit of quantum optimization often envisions harnessing uniquely quantum phenomena; however, recent work from Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, and Quantinuum demonstrates a surprising path toward competitive performance using entirely classical computation. Researchers have successfully implemented Pauli Correlation Encoding (PCE) with a focus on Instantaneous Quantum Polynomial (IQP) circuits, achieving solutions for optimization problems ranging from tens to thousands of variables. These methods produce high-quality solutions across benchmarks.

This broad applicability suggests the method isn’t limited to a specific niche within the optimization landscape. The researchers, Daniele Lizzio Bosco, Gabriel Matos, Chen-Yu Liu, Frederic Rapp, Fabian Finger, Enrico Rinaldi, Konstantinos Meichanetzidis, and Quantinuum, emphasize that this work positions efficiently simulable PCE as a useful point of comparison for future quantum implementations, helping to “separate the contribution of the PCE framework itself from the additional expressive power and measurement cost of a more general quantum ansatz.” Ultimately, the goal is to identify scenarios where genuinely quantum circuit families outperform their classical counterparts within the PCE framework.

Demonstrating a classically-simulable variant of Pauli Correlation Encoding, researchers at Quantinuum have extended the reach of correlation-based optimization, a development reshaping how quantum optimization algorithms are benchmarked. While quantum computing strives for solutions intractable for conventional machines, this work reveals a surprising capability: producing high-quality solutions across problem sizes ranging from tens to thousands of variables without needing quantum hardware at all. By separating the contribution of the PCE framework from the complexities of quantum circuits, researchers can pinpoint where genuine quantum speedup emerges.

Demonstrating that a quantum-inspired approach can be fully simulated using classical computers, researchers including those at Quantinuum are pioneering a surprising turn in quantum optimization. This scalability stems from a shift in focus; instead of relying on inherently quantum correlations, the researchers constructed instances where all necessary expectation values can be computed efficiently on conventional hardware. These methods produce high-quality solutions across benchmarks for problem sizes ranging from tens to thousands of variables.