Abstract
Resistance switching, or Resistive RAM (RRAM) devices show considerable potential for application in hardware spiking neural networks (neuro-inspired computing) by mimicking some of the behavior of biological synapses, and hence enabling non-von Neumann computer architectures. Spike-timing dependent plasticity (STDP) is one such behavior, and one example of several classes of plasticity that are being examined with the aim of finding suitable algorithms for application in many computing tasks such as coincidence detection, classification and image recognition. In previous work we have demonstrated that the neuromorphic capabilities of silicon-rich silicon oxide (SiOx) resistance switching devices extend beyond plasticity to include thresholding, spiking, and integration. We previously demonstrated such behaviors in devices operated in the unipolar mode, opening up the question of whether we could add plasticity to the list of features exhibited by our devices. Here we demonstrate clear STDP in unipolar devices. Significantly, we show that the response of our devices is broadly similar to that of biological synapses. This work further reinforces the potential of simple two-terminal RRAM devices to mimic neuronal functionality in hardware spiking neural networks.
Highlights
Non-von Neumann computing architecture, inspired by the neuronal architecture of the brain, is receiving considerable interest as a more efficient way to tackle a range of machine learning tasks such as classification, sorting, and image recognition (Indiveri et al, 2006; Smith, 2006; Izhikevich and Edelman, 2008; Ananthanarayanan et al, 2009; Jo et al, 2010)
Details of resistance switching in intrinsic oxide Resistive RAM (RRAM) devices may be found elsewhere (Mehonic et al, 2012a,b, 2015, 2016, 2017) but to briefly summarize, application of an initial electroforming field, limited to prevent catastrophic dielectric breakdown, across a pristine, highly insulating, oxide generates a filament of conductive oxygen vacancies bridging the oxide film
We have shown that it is possible to achieve an Spike-timing dependent plasticity (STDP) response from unipolar silicon oxide (SiOx) RRAM devices
Summary
Non-von Neumann computing architecture, inspired by the neuronal architecture of the brain, is receiving considerable interest as a more efficient way to tackle a range of machine learning tasks such as classification, sorting, and image recognition (Indiveri et al, 2006; Smith, 2006; Izhikevich and Edelman, 2008; Ananthanarayanan et al, 2009; Jo et al, 2010). While much work remains to be done in this field, the reduction of CMOS circuitry footprint in neuromorphic systems could yield savings in valuable silicon chip real-estate. Other benefits such as power reduction are more uncertain, and remain so until further work has been performed on algorithm development and a fuller understanding of how devices can be designed to operate at very low current and/or duty cycle. Building on our previous work, here we report, for the first time, spike timing dependent plasticity in such unipolar devices—a result that, when combined with neuronal functionality, promises a greatly reduced component count for RRAM hardware spiking neural networks. This study serves as a proof of principle of device functionality; energy efficiency will be studied in our future work
Sign-in/Register to view highlights & summary 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 callDisclaimer: 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.
