Indian Journal of Science and Technology
DOI: 10.17485/ijst/2018/v11i40/132289
Year: 2018, Volume: 11, Issue: 40, Pages: 1-6
Original Article
Savarimuthu Prakash1 * and Sam Thomas2
1 Department of Electronics and Communication Engineering, Saveetha Institute of Medical and Technical Sciences, Velappanchavadi, Chennai - 600077, Tamil Nadu, India; [email protected]
2 Information and Communication Engineering, Anna University, Guindy, Chennai - 600025, Tamil Nadu, India; [email protected]
*Author for correspondence
Savarimuthu Prakash,
Department of Electronics and Communication Engineering, Saveetha Institute of Medical and Technical Sciences, Velappanchavadi, Chennai - 600077, Tamil Nadu, India; [email protected]
Objectives: In this study, a TiO2 memristor with a non-linear ion drift model is analytically estimated for the power reduction in a basic analog circuit. Methods/Analysis: For the estimation of power reduction, the non-inverting and inverting amplifier configurations for conventional resistance and memristors are considered. The circuit is chosen since the output is directly proportional to the input and the component values. The recent progress in the experimental realization of memristive devices has renewed interest in artificial neural networks. Findings: To emulate human brain like functionalities at circuit level require two components, neurons and the connecting synapses are required in artificial neural networks. The synapse is a crucial element in biological neural networks. The memristor has been predicated as electronic equivalent of biological synapse. Basically memristor, a resistor with memory; non-volatile and its response depends on continuous set of resistance values, making it ideal for tuning synaptic weights of neuromorphic cells. The first observation on analytical estimate power reduction from conventional op-amp based non-inverting and inverting amplifier circuits to that of memristance based op-amp non-inverting and inverting amplifier circuits indicate 99% reduction in power consumption. Secondly, by varying the amplitude of the input voltage resulted in varying power dissipation in conventional amplifier but resistance values remains constant as expected. However, by varying the amplitude of the input voltage applied to the memristor the power dissipation in the circuit provide an empirical estimating result a clear variation in memristance. Novelty/Improvement: Hence, this phenomenon indicates that weighted resistance function in a synapse can be implemented using memristors. The feasible scaling-up to approach real device densities requires reduced power consumption.
Keywords: Device Density, Inverting Amplifier, Memristance, Neuromorphic Cell, Non-inverting Amplifier, Non-linear Ion Drift Model
Subscribe now for latest articles and news.
