Neuromorphic Low-Power Inference on Memristive Crossbars With On-Chip Offset Calibration

Abstract

Monolithic integration of silicon with nano-sized Redox-based resistive Random-Access Memory (ReRAM) devices opened the door to the creation of dense synaptic connections for bio-inspired neuromorphic circuits. One drawback of OxRAM based neuromorphic systems is the relatively low ON resistance of OxRAM synapses (in the range of just a few kilo-ohms). This requires relatively large currents (many micro amperes per synapse), and therefore imposes strong driving capability demands on peripheral circuitry, limiting scalability and low power operation. After learning, however, a read inference can be made low-power by applying very small amplitude read pulses, which require much smaller driving currents per synapse. Here we propose and experimentally demonstrate a technique to reduce the amplitude of read inference pulses in monolithic neuromorphic CMOS OxRAM-synaptic crossbar systems. Unfortunately, applying tiny read pulses is non-trivial due to the presence of random DC offset voltages. To overcome this, we propose finely calibrating DC offset voltages using a bulk-based three-stage on-chip calibration technique. In this work, we demonstrate spiking pattern recognition using STDP learning on a small 4x4 proof-of-concept memristive crossbar, where on-chip offset calibration is implemented and inference pulse amplitude could be made as small as 2mV. A chip with pre-synaptic calibrated input neuron drivers and a 4x4 1T1R synapse crossbar was designed and fabricated in the CEA-LETI MAD200 technology, which uses monolithic integration of OxRAMs above ST130nm CMOS. Custom-made PCBs hosting the postsynaptic circuits and control FPGAs were used to test the chip in different experiments, including synapse characterization, template matching, and pattern recognition using STDP learning, and to demonstrate the use of on-chip offset-calibrated low-power amplifiers. According to our experiments, the minimum possible inference pulse amplitude is limited by offset voltage drifts and noise. We conclude the paper with some suggestions for future work in this direction.

Comprehensive Model of Electron Conduction in Oxide-Based Memristive Devices

“…One needs to consider, however, that the description of the different conduction states needs to be consistent with the idea of the reconfiguration of the oxygen defects. It should be noted that in state-of-the-art VCM cells, which show reproducible switching with high speed at comparably low voltages, the oxide layer is very thin and asymmetric electrode materials are used. − …”

ASM-ROBOT: A Cyber-Physical Home Automation Controller with Memristive Reconfigurable State Machine

“…This work has similar baseline design attributes with robotic smart homes based on stabilized feedback Episodic memory (SF-EM) [58], stewards robot smart homes [59], humanoid defense smart homes [60], Indoor autonomous robots [61], and Smart home activities IoT [62]. These similarities are highlighted below [63], [64], [65]:…”

“…Considering the memristive neuromorphic circuits used for computation acceleration, the work [42], [43], [44], showed that the synaptic circuit is required to represent process variable weights in binary form. The works equally highlighted that various synaptic circuits can be designed to utilize neuromorphic circuits.…”

Uniform self-rectifying resistive random-access memory based on an MXene-TiO2 Schottky junction

“…In addition, the current of our devices with the smallest resistive area is more than 2 nA, which is big enough to read. In the pratical chip design, even the current of fA level can be collected by adding auxiliary circuits, such as operational amplifiers [41][42][43] .…”