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

Abstract

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. Multiphysical 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.

Emulation of spike-timing dependent plasticity in nano-scale phase change memory

“…Suri et al [15] introduced two-phase change memory cells that can play specific roles in flowing LTP and LTD currents, but they require a large area for cell-selecting transistors and extra current comparator circuits. Eryilmaz et al [16] presented a 10 Â 10 array of phase change synaptic devices that showed simple pattern learning and recognition abilities, but they had to use one metal-oxide-semiconductor (MOS) switch per synaptic device.…”

“…Recently, single resistive memory devices with transition metal oxides [13], Ge-Te-Sb phase change material [14][15][16], and organic material [17] as electronic synapses have gathered attention, because they can be operated with low power consumption. They have a simple structure that can be easily integrated to have a high density that is comparable to the number of neurons in our neocortex ( $10 11 ) by using nano fabrication techniques in the semiconductor industry.…”

Trade-off between number of conductance states and variability of conductance change in Pr0.7Ca0.3MnO3-based synapse device

“…[10][11][12][13][15][16][17] Even though each synapse device has its advantages, the synapse device needs a simple two-terminal structure for high density, low-power operation, an analogous conductance change, and reliable characteristics for the implementation of the biological-brain-like S-HNN. 5 In these respects, ReRAM can be a promising candidate.…”

Probabilistic neuromorphic system using binary phase-change memory (PCM) synapses: Detailed power consumption analysis

“…This study was carried out to simplify system programming and optimize the synaptic power consumption. Binary PCM programming is easier compared to multilevel operation and does not require specific "refresh" operations [4]. We present 1 programming schemes for both matrix structures with selector device (1T-1R) and selector-free (1R) crossbar architectures.…”

“…PCM-based neuromorphic systems proposed until now follow a multilevel-deterministic approach [2], [4], where the weight of the PCM synapses, i.e. the strength of the connections between neurons, is gradually tuned among several levels between a minimum and a maximum value.…”