We attempted to replicate the changes in the membrane potential of neurons using thin-film neuromorphic devices that integrate memristors and high-resistance SiO2. Using the dynamic behavior of memristors circuit, we constructed the sum-of-products circuit for continuous spike pulses and confirmed its operation. We confirmed the staircase waveform that the membrane potential rises while spike pulses are applied and is maintained while no spike pulses are applied. We confirmed that this waveform changes depending on the resistance value of memristors. Therefore, we thought that the fabricated devices could be used as synapses in neural networks.