Pim van den Heuvel and colleagues have developed three new QUBO formulations designed to reduce slack variables when solving the Knapsack problem using quantum computers. The team’s work addresses a common challenge in formulating optimization problems for quantum solvers, potentially improving performance across different platforms. One new method avoids slack variables altogether. Both simulations using the linear ramp Quantum Approximate Optimization Algorithm and tests on a D-Wave quantum annealer confirmed improvements in finding optimal solutions with these new approaches.

QUBO Formulations Reduce Slack Variables in Knapsack Problem

The most substantial gains in solving the Knapsack problem using quantum computing came from a QUBO formulation that eliminated slack variables, increasing success probabilities by several orders of magnitude even when tackling relatively small problem instances. Researchers Pim van den Heuvel, Vrinda Mehta, Dennis Willsch, J. A. Montanez-Barrera, and Kristel Michielsen detailed these three QUBO formulations in a paper submitted on November 4, 2026.

These modifications aim to minimize the resources required by QUBO solvers, a critical factor in practical applications of quantum optimization. The authors wrote that all three approaches increase the probability of finding the optimal solution, highlighting the broad applicability of their findings.

👉 More information
🗞 Resource-efficient formulations of the Knapsack problem for quantum computing
✍️ Pim van den Heuvel, Vrinda Mehta, Dennis Willsch, J. A. Montanez-Barrera and Kristel Michielsen
🧠 ArXiv: https://arxiv.org/abs/2611.06797

More like this