Scientists have developed a new type of probabilistic computer that uses a hypergraph-based approach to solve Boolean Satisfiability (SAT) problems. This system, which is neither a quantum computer nor a quantum annealer, operates by the interaction of optical parametric oscillators to explore the solution space of SAT problems, a class of NP-complete combinatorial optimization problems fundamental in computer science and various scientific fields. The novelty lies in its architecture, which allows encoding the logical relationships of SAT problems into the connectivity and interactions of the oscillators.

The proposed method represents Boolean variables and clauses of the SAT problem as nodes and edges of a hypergraph, where edges connect multiple nodes. Each optical parametric oscillator, which can exist in two stable phases (0 or π), represents a Boolean variable. Interactions between these oscillators are designed to favor configurations that satisfy the problem's clauses. By allowing the system to evolve probabilistically, it seeks the minimum energy state that corresponds to a solution to the SAT problem. This approach differs from conventional digital computers, which explore solutions sequentially, and from quantum computers, which use superposition and entanglement.

Experimental results demonstrate that this probabilistic computer can efficiently solve moderately sized SAT problems, showing promising scalability. Although it does not currently surpass the capability of the most advanced classical algorithms for very large SAT problems, its architecture and operating principles open a new avenue for unconventional computing. The ability to encode complex logical problems into the dynamics of physical systems offers an alternative path to address difficult computational challenges, with potential implications in areas such as artificial intelligence, optimization, and cryptography. Future improvements in implementation and the number of oscillators are expected to allow tackling problems of greater complexity.