📝 SummarXiv

Abstract

Many combinatorial optimization problems can be reformulated as the task of finding the ground state of a physical system, such as the Ising model. Most existing Ising solvers are inspired by simulated annealing. Although annealing techniques offer scalability, they lack convergence guarantees and are sensitive to the cooling schedule. We propose to solve the Ising problem by relaxing the binary spins to continuous variables and introducing a potential function (attractor) that steers the solution toward binary spin configurations. The resulting Hamiltonian can be expressed as a difference of convex functions, enabling the design of efficient iterative algorithms that require a single matrix-vector multiplication per iteration and are backed by convergence guarantees. We implement our Ising solver across a range of GPU platforms: from edge devices to high-performance computing clusters and demonstrate that it consistently outperforms existing solvers across problem sizes ranging from small ($10^3$ spins) to ultra-large ($10^8$ spins).