Abstract:The progress in neuromorphic computing is fueled by the development of novel nonvolatile memories capable of storing analog information and implementing neural computation efficiently. However, like most other analog circuits, these devices and circuits are prone to imperfections, such as temperature dependency, noise, tuning error, etc., often leading to considerable performance degradation in neural network implementations. Indeed, imperfections are major obstacles in the path of further progress and ultimate commercialization of these technologies. Hence, a practically viable approach should be developed to deal with these nonidealities and unleash the full potential of nonvolatile memories in neuromorphic systems. Here, for the first time, we report a comprehensive characterization of critical imperfections in two analog-grade memories, namely passively-integrated memristors and redesigned eFlash memories, which both feature long-term retention, high endurance, analog storage, low-power operation, and compact nano-scale footprint. Then, we propose a holistic approach that includes modifications in the training, tuning algorithm, memory state optimization, and circuit design to mitigate these imperfections. Our proposed methodology is corroborated on a hybrid software/experimental framework using two benchmarks: a moderate-size convolutional neural network and ResNet-18 trained on CIFAR-10 and ImageNet datasets, respectively. Our proposed approaches allow 2.5x to 9x improvements in the energy consumption of memory arrays during inference and sub-percent accuracy drop across 25-100 C temperature range. The defect tolerance is improved by >100x, and a sub-percent accuracy drop is demonstrated in deep neural networks built with 64x64 passive memristive crossbars featuring 25% normalized switching threshold variations.

Mitigating the Effect of Reliability Soft-errors of RRAM Devices on the Performance of RRAM-based Neuromorphic Systems

Analysis of RRAM Reliability Soft-Errors on the Performance of RRAM-Based Neuromorphic Systems.

A Study of the Effect of RRAM Reliability Soft Errors on the Performance of RRAM-Based Neuromorphic Systems

Robustness of Neuromorphic Computing with RRAM-based Crossbars and Optical Neural Networks.

Reliable Memristor-based Neuromorphic Design Using Variation- and Defect-Aware Training

Impact of the Interconnection Structure on RRAM-Based Crossbar Array in Neuromorphic Chip

A High-Speed and High-Efficiency Diverse Error Margin Write-Verify Scheme for an RRAM-Based Neuromorphic Hardware Accelerator

Thermal Induced Retention Degradation of RRAM-based Neuromorphic Computing Chips

Circuit Modeling for RRAM-Based Neuromorphic Chip Crossbar Array with and Without Write-Verify Scheme.

Reduction and IR-drop Compensations Techniques for Reliable Neuromorphic Computing Systems

Mitigating Imperfections in Mixed-Signal Neuromorphic Circuits

Understanding the trade-offs of device, circuit and application in ReRAM-based neuromorphic computing systems

A Novel Neuromorphic Hardware Using Area-Efficient Chain RRAM-Based Synapses and Compact Neurons with (Anti-) Integration Scheme

On the Mitigation of Read Disturbances in Neuromorphic Inference Hardware

Reliability Perspective on Neuromorphic Computing Based on Analog RRAM

Rescuing RRAM-Based Computing from Static and Dynamic Faults

Device and circuit optimization of RRAM for neuromorphic computing

A Compact Model of Analog RRAM With Device and Array Nonideal Effects for Neuromorphic Systems

Compensation Architecture to Alleviate Noise Effects in RRAM-based Computing-in-memory Chips with Residual Resource

Improving Reliability of Non-Volatile Memory Technologies Through Circuit Level Techniques and Error Control Coding

Fault Tolerance in Neuromorphic Computing Systems