[16p-D215-1~16] 11.5 Junction and circuit fabrication process, digital applications

△ [16p-D215-13] Stabilization of Boltzmann machines by superconducting circuits

〇Seigo Ito¹, Yuki Yamanashi^1,2, Nobuyuki Yoshikawa^1,2 (1.Yokohama Nat'l Univ., 2.IAS,Yokohama Nat'l Univ.)

Keywords:Boltzmann machines, Quantum flux parametrons

A Boltzmann machine is an artificial neural network. The output is obtained by stochastic transition of the entire system to an energy stable point. We designed a NOR as a logic gate of the Boltzmann machine using a quantum flux parametron, which is one of superconducting circuits. We used maximum likelihood estimation to optimize the operating point, which is determined by the bias current and mutual inductance supplied to each QFP.