TY - GEN A1 - Milo, Valerio A1 - Anzalone, Francesco A1 - Zambelli, Cristian A1 - Pérez, Eduardo A1 - Mahadevaiah, Mamathamba Kalishettyhalli A1 - Ossorio, Óscar G. A1 - Olivo, Piero A1 - Wenger, Christian A1 - Ielmini, Daniele T1 - Optimized programming algorithms for multilevel RRAM in hardware neural networks T2 - IEEE International Reliability Physics Symposium (IRPS), 2021 N2 - A key requirement for RRAM in neural network accelerators with a large number of synaptic parameters is the multilevel programming. This is hindered by resistance imprecision due to cycle-to-cycle and device-to-device variations. Here, we compare two multilevel programming algorithms to minimize resistance variations in a 4-kbit array of HfO 2 RRAM. We show that gate-based algorithms have the highest reliability. The optimized scheme is used to implement a neural network with 9-level weights, achieving 91.5% (vs. software 93.27%) in MNIST recognition. KW - RRAM KW - Multilevel switching KW - neural network KW - memristive switching Y1 - 2021 SN - 978-1-7281-6894-4 U6 - https://doi.org/10.1109/IRPS46558.2021.9405119 SN - 1938-1891 ER - TY - GEN A1 - Romero-Zaliz, Rocío A1 - Pérez, Eduardo A1 - Jimenez-Molinos, Francisco A1 - Wenger, Christian A1 - Roldan, Juan Bautista T1 - Influence of variability on the performance of HfO2 memristor-based convolutional neural networks T2 - Solid State Electronics N2 - A study of convolutional neural networks (CNNs) was performed to analyze the influence of quantization and variability in the network synaptic weights. Different CNNs were considered accounting for the number of convolutional layers, size of the filters in the convolutional layer, number of neurons in the final network layers and different sets of quantization levels. The conductance levels of fabricated 1T1R structures based on HfO2 memristors were considered as reference for four or eight level quantization processes at the inference stage of the CNNs, which were previous trained with the MNIST dataset. We also included the variability of the experimental conductance levels that was found to be Gaussian distributed and was correspondingly modeled for the synaptic weight implementation. KW - RRAM KW - neural network KW - HfO2 KW - memristive switching Y1 - 2021 U6 - https://doi.org/10.1016/j.sse.2021.108064 SN - 0038-1101 VL - 185 ER - TY - GEN A1 - Pérez, Eduardo A1 - Mahadevaiah, Mamathamba Kalishettyhalli A1 - Perez-Bosch Quesada, Emilio A1 - Wenger, Christian T1 - In-depth characterization of switching dynamics in amorphous HfO2 memristive arrays for the implementation of synaptic updating rules T2 - Japanese Journal of Applied Physics N2 - Accomplishing truly analog conductance modulation in memristive arrays is crucial in order to implement the synaptic plasticity in hardware-based neuromorphic systems. In this paper, such a feature was addressed by exploiting the inherent stochasticity of switching dynamics in amorphous HfO2 technology. A thorough statistical analysis of experimental characteristics measured in 4 kbit arrays by using trains of identical depression/potentiation pulses with different voltage amplitudes and pulse widths provided the key to develop two different updating rules and to define their optimal programming parameters. The first rule is based on applying a specific number of identical pulses until the conductance value achieves the desired level. The second one utilized only one single pulse with a particular amplitude to achieve the targeted conductance level. In addition, all the results provided by the statistical analysis performed may play an important role in understanding better the switching behavior of this particular technology. KW - RRAM KW - memristive device KW - HfO2 KW - memristive switching Y1 - 2022 U6 - https://doi.org/10.35848/1347-4065/ac6a3b SN - 0021-4922 VL - 61 SP - 1 EP - 7 ER - TY - GEN A1 - Glukhov, Artem A1 - Lepri, Nicola A1 - Milo, Valerio A1 - Baroni, Andrea A1 - Zambelli, Cristian A1 - Olivo, Piero A1 - Pérez, Eduardo A1 - Wenger, Christian A1 - Ielmini, Daniele T1 - End-to-end modeling of variability-aware neural networks based on resistive-switching memory arrays T2 - Proc. 30th IFIP/IEEE International Conference on Very Large Scale Integration (VLSI-SoC 2022) N2 - Resistive-switching random access memory (RRAM) is a promising technology that enables advanced applications in the field of in-memory computing (IMC). By operating the memory array in the analogue domain, RRAM-based IMC architectures can dramatically improve the energy efficiency of deep neural networks (DNNs). However, achieving a high inference accuracy is challenged by significant variation of RRAM conductance levels, which can be compensated by (i) advanced programming techniques and (ii) variability-aware training (VAT) algorithms. In both cases, however, detailed knowledge and accurate physics-based statistical models of RRAM are needed to develop programming and VAT methodologies. This work presents an end-to-end approach to the development of highly-accurate IMC circuits with RRAM, encompassing the device modeling, the precise programming algorithm, and the VAT simulations to maximize the DNN classification accuracy in presence of conductance variations. KW - RRAM KW - HfO2 KW - neural network KW - memristive switching Y1 - 2022 U6 - https://doi.org/10.1109/VLSI-SoC54400.2022.9939653 SP - 1 EP - 5 ER -