Biologically Plausible Boltzmann Machine

The dichotomy in power consumption between digital and biological information processing systems is an intriguing open question related at its core with the necessity for a more thorough understanding of thermodynamics logic computing. To contribute this regard, we put forward model that implements Boltzmann machine (BM) approach to computation through electric substrate under thermal fluctuations dissipation. resulting network has precisely defined statistical properties, which are consistent data accessible BM. It shown by proposed model, it possible design neural-inspired gates capable universal Turing similar conditions those found neural networks storage potentials comparable scales.