Proposed modified computational model for the amoeba-inspired combinatorial optimization machine

In the modern information society, conventional Neumann-type computers face a serious problem of increased complexity and cost in computation. While the information processing capacity of computers is predicted to reach its limits because of the limit of Moore's law and the von Neumann bottleneck, the required computing power is increasing exponentially. Under these circumstances, new efficient domain-specific computational architectures beyond the von Neumann type are demanded [1]. The serious problem particularly appears in combinatorial optimization problems [2]. Recently, the Ising machine has been proposed as a solution and its practical applications are being investigated [3-7]. There, the combinatorial optimization problem is mapped onto the ground-state search of the Ising model, which is originally a mathematical model of magnetic materials [8, 9]. On the other hand, research has been intensively conducted to construct combinatorial optimization machines that mimic the information processing of living organisms. This is because organisms often perform sophisticated information processing such as image recognitions, sound recognitions, and optimizations with low energy consumption. One of the simplest and most useful machines of such kinds is a computing technology inspired by the information processing of the amoeboid organism [10, 11].