Abstract
Hardware processors for neuromorphic computing are gaining significant interest as they offer the possibility of real in-memory computing, thus by-passing the limitations of speed and energy consumption of the von Neumann architecture. One of the major limitations of current neuromorphic technology is the lack of bio-realistic and scalable devices to improve the current design of artificial synapses and neurons. To overcome these limitations, the emerging technology of resistive switching memory has attracted wide interest as a nano-scaled synaptic element. This paper describes the implementation of a perceptron-like neuromorphic hardware capable of spike-timing dependent plasticity (STDP), and its operation under stochastic learning conditions. The learning algorithm of a single or multiple patterns, consisting of either static or dynamic visual input data, is described. The impact of noise is studied with respect to learning efficiency (false fire, true fire) and learning time. Finally, the impact of stochastic learning rule, such as the inversion of the time dependence of potentiation and depression in STDP, is considered. Overall, the work provides a proof of concept for unsupervised learning by STDP in memristive networks, providing insight into the dynamics of stochastic learning and supporting the understanding and design of neuromorphic networks with emerging memory devices.
Highlights
N EUROMORPHIC circuits are gaining momentum as a new computing platform for brain-inspired learning and inference with widespread applications including robotics, Manuscript received June 14, 2017; revised October 6, 2017; accepted November 8, 2017
Of learning, which can be explained based on the anti-spiketiming dependent plasticity (STDP) shape in Fig. 11(b): as soon as pattern synapses are randomly potentiated, their enhanced weights increase the probability of fire in response to the presentation of a pattern as input data
We presented a neuromorphic hardware with RRAM synapses which is capable of unsupervised learning by STDP
Summary
N EUROMORPHIC circuits are gaining momentum as a new computing platform for brain-inspired learning and inference with widespread applications including robotics, Manuscript received June 14, 2017; revised October 6, 2017; accepted November 8, 2017. Information is generally carried by spikes: neurons integrate incoming input spikes and generate an output spike as the internal potential exceeds a certain firing threshold [2] Such event-driven computing scheme is at the basis of the high energy efficiency of neuromorphic hardware, as energy is consumed only when and where information is processed [3]. Spike-timing dependent plasticity (STDP), one of the weight update processes in biological neural networks, was demonstrated in both PCM [10], [11] and RRAM [12]–[15] by overlapping pulses at 2 terminals of the memory devices. Brain inspired circuits generally consist of reconfigurable networks of spiking neurons and plastic synapses, where STDP is a key learning scheme [19]. Controlling and optimizing STDP-based unsupervised learning in hybrid synapses are essential steps toward the development of large-scale spiking neural networks capable of plasticity
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IEEE Journal on Emerging and Selected Topics in Circuits and Systems
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
