Damien Querlioz

Bio-Inspired Stochastic Computing Using Binary CBRAM Synapses

In this paper, we present an alternative approach to neuromorphic systems based on multilevel resistive memory synapses and deterministic learning rules. We demonstrate an original methodology to use conductive-bridge RAM (CBRAM) devices as, easy to program and low-power, binary synapses with stochastic learning rules. New circuit architecture, programming strategy, and probabilistic spike-timing dependent plasticity (STDP) learning rule for two different CBRAM configurations with-selector (1T-1R) and without-selector (1R) are proposed. We show two methods (intrinsic and extrinsic) for implementing probabilistic STDP rules. Fully unsupervised learning with binary synapses is illustrated through two example applications: 1) real-time auditory pattern extraction (inspired from a 64-channel silicon cochlea emulator); and 2) visual pattern extraction (inspired from the processing inside visual cortex). High accuracy (audio pattern sensitivity > 2, video detection rate > 95%) and low synaptic-power dissipation (audio 0.55 μW, video 74.2 μW) are shown. The robustness and impact of synaptic parameter variability on system performance are also analyzed.

Phase change memory as synapse for ultra-dense neuromorphic systems: Application to complex visual pattern extraction

We demonstrate a unique energy efficient methodology to use Phase Change Memory (PCM) as synapse in ultra-dense large scale neuromorphic systems. PCM devices with different chalcogenide materials were characterized to demonstrate synaptic behavior. Multi-physical simulations were used to interpret the results. We propose special circuit architecture (“the 2-PCM synapse”), read, write, and reset programming schemes suitable for the use of PCM in neural networks. A versatile behavioral model of PCM which can be used for simulating large scale neural systems is introduced. First demonstration of complex visual pattern extraction from real world data using PCM synapses in a 2-layer spiking neural network (SNN) is shown. System power analysis for different scaled PCM technologies is also provided.

Immunity to Device Variations in a Spiking Neural Network With Memristive Nanodevices

Memristive nanodevices can feature a compact multilevel nonvolatile memory function, but are prone to device variability. We propose a novel neural network-based computing paradigm, which exploits their specific physics, and which has virtual immunity to their variability. Memristive devices are used as synapses in a spiking neural network performing unsupervised learning. They learn using a simplified and customized “spike timing dependent plasticity” rule. In the network, neurons’ threshold is adjusted following a homeostasis-type rule. We perform system level simulations with an experimentally verified model of the memristive devices’ behavior. They show, on the textbook case of character recognition, that performance can compare with traditional supervised networks of similar complexity. They also show that the system can retain functionality with extreme variations of various memristive devices’ parameters (a relative standard dispersion of more than 50% is tolerated on all device parameters), thanks to the robustness of the scheme, its unsupervised nature, and the capability of homeostasis. Additionally the network can adjust to stimuli presented with different coding schemes, is particularly robust to read disturb effects and does not require unrealistic control on the devices’ conductance. These results open the way for a novel design approach for ultraadaptive electronic systems.

CBRAM devices as binary synapses for low-power stochastic neuromorphic systems: Auditory (Cochlea) and visual (Retina) cognitive processing applications

In this work, we demonstrate an original methodology to use Conductive-Bridge RAM (CBRAM) devices as binary synapses in low-power stochastic neuromorphic systems. A new circuit architecture, programming strategy and probabilistic STDP learning rule are proposed. We show, for the first time, how the intrinsic CBRAM device switching probability at ultra-low power can be exploited to implement probabilistic learning rule. Two complex applications are demonstrated: real-time auditory (from 64-channel human cochlea) and visual (from mammalian visual cortex) pattern extraction. A high accuracy (audio pattern sensitivity >2, video detection rate >95%) and ultra-low synaptic-power dissipation (audio 0.55μW, video 74.2μW) are obtained.

Plasticity in memristive devices for spiking neural networks.

Memristive devices present a new device technology allowing for the realization of compact non-volatile memories. Some of them are already in the process of industrialization. Additionally, they exhibit complex multilevel and plastic behaviors, which make them good candidates for the implementation of artificial synapses in neuromorphic engineering. However, memristive effects rely on diverse physical mechanisms, and their plastic behaviors differ strongly from one technology to another. Here, we present measurements performed on different memristive devices and the opportunities that they provide. We show that they can be used to implement different learning rules whose properties emerge directly from device physics: real time or accelerated operation, deterministic or stochastic behavior, long term or short term plasticity. We then discuss how such devices might be integrated into a complete architecture. These results highlight that there is no unique way to exploit memristive devices in neuromorphic systems. Understanding and embracing device physics is the key for their optimal use.

Suppression of the orientation effects on bandgap in graphene nanoribbons in the presence of edge disorder

This letter shows that a moderate degree of edge disorder can explain the fact that the experimentally measured bandgaps of graphene nanoribbons (GNRs) do not depend on orientation. We argue that GNRs actually behave similarly to Anderson insulators and the measured bandgaps should thus be interpreted as quasi-mobility edges. Calculations in the tight binding approach reveal that in the presence of edge disorder, quasi-mobility edge and electronic structures become independent of orientation and that quasi-mobility edge follows a quasi-universal law similar to experimental data, although with different parameters.

Visual Pattern Extraction Using Energy-Efficient “2-PCM Synapse” Neuromorphic Architecture

We introduce a novel energy-efficient methodology “2-PCM Synapse” to use phase-change memory (PCM) as synapses in large-scale neuromorphic systems. Our spiking neural network architecture exploits the gradual crystallization behavior of PCM devices for emulating both synaptic potentiation and synaptic depression. Unlike earlier attempts to implement a biological-like spike-timing-dependent plasticity learning rule with PCM, we use a simplified rule where long-term potentiation and long-term depression can both be produced with a single invariant crystallizing pulse. Our architecture is simulated on a special purpose event-based simulator, using a behavioral model for the PCM devices validated with electrical characterization. The system, comprising about 2 million synapses, directly learns from event-based dynamic vision sensors. When tested with real-life data, it is able to extract complex and overlapping temporally correlated features such as car trajectories on a freeway. Complete trajectories can be learned with a detection rate above 90 %. The synaptic programming power consumption of the system during learning is estimated and could be as low as 100 nW for scaled down PCM technology. Robustness to device variability is also evidenced.

2012 Special Issue: Extraction of temporally correlated features from dynamic vision sensors with spike-timing-dependent plasticity

A biologically inspired approach to learning temporally correlated patterns from a spiking silicon retina is presented. Spikes are generated from the retina in response to relative changes in illumination at the pixel level and transmitted to a feed-forward spiking neural network. Neurons become sensitive to patterns of pixels with correlated activation times, in a fully unsupervised scheme. This is achieved using a special form of Spike-Timing-Dependent Plasticity which depresses synapses that did not recently contribute to the post-synaptic spike activation, regardless of their activation time. Competitive learning is implemented with lateral inhibition. When tested with real-life data, the system is able to extract complex and overlapping temporally correlated features such as car trajectories on a freeway, after only 10 min of traffic learning. Complete trajectories can be learned with a 98% detection rate using a second layer, still with unsupervised learning, and the system may be used as a car counter. The proposed neural network is extremely robust to noise and it can tolerate a high degree of synaptic and neuronal variability with little impact on performance. Such results show that a simple biologically inspired unsupervised learning scheme is capable of generating selectivity to complex meaningful events on the basis of relatively little sensory experience.

Robust neural logic block (NLB) based on memristor crossbar array

Neural networks are considered as promising candidates for implementing functions in memristor crossbar array with high tolerance to device defects and variations. Based on such arrays, Neural Logic Blocks (NLB) with learning capability can be built to replace Configurable Logic Block (CLB) in programmable logic circuits. In this article, we describe a neural learning method to implement Boolean functions in memristor NLB. By using Monte-Carlo simulation, we demonstrate its high robustness against most important device defects and variations, like static defects and memristor voltage threshold variability.

On the Ability of the Particle Monte Carlo Technique to Include Quantum Effects in Nano-MOSFET Simulation

In this paper, we report on the possibility of using particle-based Monte Carlo (MC) techniques to incorporate all relevant quantum effects in the simulation of semiconductor nanotransistors. Starting from the conventional MC approach within the semiclassical Boltzmann approximation, we develop a multisubband description of transport to include quantization in ultrathin-body devices. This technique is then extended to the particle simulation of quantum transport within the Wigner formulation. This new simulator includes all expected quantum effects in nanotransistors and all relevant scattering mechanisms, which are taken into account the same way as in Boltzmann simulation. This paper is illustrated by analyzing the device operation and performance of multigate nanotransistors in a convenient range of channel lengths and thicknesses to separate the influence of all relevant effects: Significant quantization effects occur for thickness smaller than 5 nm and wave-mechanical-transport effects manifest themselves for channel length smaller than 10 nm. We also show that scattering mechanisms still have an important influence in nanoscaled double-gate transistors, both in the intrinsic part of the channel and in the resistive lateral extensions.