Implementation of High-Speed Single Flux-Quantum Up/Down Counter for the Neural ComputationUsing Stochastic Logic

We report a design and an experimental result of an SFQ up/down counter for neural computation using stochastic logic. Neural computation using stochastic logic must accumulate the pulses in order to generate the membrane potential of a neuron. A high-speed up/down counter is necessary to achieve high-speed operation. The proposed up/down counter has two operation phases which are the accumulation of signals and the access of accumulation result. Even if the number of bit increases, the operation speed of the accumulation does not decrease in this method. A 4-bit up/down counter is fabricated using niobium-trilayer standard process and successfully demonstrated. The numerical simulation shows that the up/down counter can operate up to 70 GHz with an enough bias margin in the accumulation phase.