Flatiron Institute Matches Quantum Spin-Glass Results With 1 Classical System
Updated
Updated · ScienceAlert · May 29
Flatiron Institute Matches Quantum Spin-Glass Results With 1 Classical System
5 articles · Updated · ScienceAlert · May 29
Flatiron Institute researchers simulated quantum spin glasses on a classical computer, reporting results as good as or better than D-Wave’s earlier quantum runs on cylindrical, diamond and cubic lattices.
New compression methods drove the advance: tensor networks trimmed redundant quantum information, while belief propagation extracted key signals cheaply enough that some calculations ran on a laptop.
Larger 3D geometries still needed high-end chips and graphics hardware, but the setup remained fully classical despite a problem long viewed as requiring quantum machines.
The result narrows where quantum computers hold a clear edge and gives researchers a classical benchmark to check and guide future quantum-computing work.
A classical computer just solved a problem meant for quantum machines. Is the era of quantum supremacy already being challenged?
If classical algorithms can now simulate quantum physics, what is the real purpose of building quantum computers?
In May 2026, the Flatiron Institute published a groundbreaking Science paper showing that advanced classical algorithms, specifically 3D tensor network methods, can simulate complex quantum systems as accurately as D-Wave's quantum annealer. This work directly challenges D-Wave's earlier claim that simulating disordered atomic-level magnets, known as spin glasses, was beyond classical computation. By addressing this core problem, the Flatiron team demonstrated that modern classical computing can match or even surpass quantum results in certain domains, raising new questions about the boundaries of quantum supremacy and pushing the field toward more rigorous benchmarks.