For the model of so-called coherent computing recently proposed by Yamamoto et al. [Y. Yamamoto et al., New Gen. Comput. 30 (2012) 327-355], a theoretical analysis of the success probability is given. Although it was claimed as their prospect that the Ising spin configuration problem would be efficiently solvable in the model, here it is shown that the probability of finding a desired spin configuration decreases exponentially in the number of spins for certain hard instances. The model is thus physically unfeasible for solving the problem within a polynomial cost.
تحميل البحث