Computational aspects of orbifold equivalence

In this paper we study the computational feasibility of an algorithm to prove orbifold equivalence between potentials describing Landau-Ginzburg models. Through a comparison with state-of-the-art results of Grobner basis computations in cryptology, we infer that the algorithm produces systems of equations that are beyond the limits of current technical capabilities. As such the algorithm needs to be augmented by 'inspired guesswork', and we provide examples of applying this approach.