Laptop tensor-network simulation matches quantum hardware

Researchers at the Simons Foundation’s Flatiron Institute and collaborators at Boston University used tensor-network methods on a standard laptop to compress the wave function of hundreds of entangled qubits and reproduce results that matched theoretical predictions and independent quantum-computer simulations. The work was published May 21, 2026, in the journal Science.

The paper, titled “Dynamics of disordered quantum systems with two- and three-dimensional tensor networks,” lists Joseph Tindall, Antonio Francesco Mello, Matthew Fishman, E. Miles Stoudenmire and Dries Sels as authors. The team combined an adapted belief-propagation–style algorithm dating to the 1980s with the ITensor software package to represent the complex many-qubit wave function in a compressed form that fits on modest hardware.

Lead author Joseph Tindall described the representation as “a zip file for the wave function.” The group did not provide exact laptop specifications in the report, emphasizing that the result rests on the mathematical representation and software rather than on a particular machine or large-scale hardware.

The laptop simulations reproduced theoretical expectations and aligned with independent simulations run on quantum hardware. The authors note that these comparisons provide a direct test of where classical tensor-network methods can perform the same tasks previously described as beyond the reach of classical machines. The paper references earlier claims associated with Andrew D. King and collaborators about the necessity of quantum hardware for this class of problems and presents contrasting results for the specific systems studied.

The researchers stated that success depended on exploiting structure and compressibility in the wave function so the classical computer need not track every amplitude individually. They also noted limitations: tensor networks compress states that have exploitable structure, and problems with higher entanglement or complexity can exceed what current compression methods handle. The paper suggests further work applying tensor-network approaches to quantum dynamics, materials research and certain optimization problems where the state can be represented compactly.

The team emphasized that the findings do not eliminate the use cases for quantum computers, and that quantum hardware remains necessary for problems whose entanglement and complexity surpass classical compression techniques.