@misc{RomeroZalizPerezJimenezMolinosetal., author = {Romero-Zaliz, Roc{\´i}o and P{\´e}rez, Eduardo and Jimenez-Molinos, Francisco and Wenger, Christian and Roldan, Juan Bautista}, title = {Study of Quantized Hardware Deep Neural Networks Based on Resistive Switching Devices, Conventional versus Convolutional Approaches}, series = {Electronics (MDPI)}, volume = {10}, journal = {Electronics (MDPI)}, number = {3}, issn = {2079-9292}, doi = {10.3390/electronics10030346}, pages = {14}, abstract = {A comprehensive analysis of two types of artificial neural networks (ANN) is performed to assess the influence of quantization on the synaptic weights. Conventional multilayer-perceptron (MLP) and convolutional neural networks (CNN) have been considered by changing their features in the training and inference contexts, such as number of levels in the quantization process, the number of hidden layers on the network topology, the number of neurons per hidden layer, the image databases, the number of convolutional layers, etc. A reference technology based on 1T1R structures with bipolar memristors including HfO2 dielectrics was employed, accounting for different multilevel schemes and the corresponding conductance quantization algorithms. The accuracy of the image recognition processes was studied in depth. This type of studies are essential prior to hardware implementation of neural networks. The obtained results support the use of CNNs for image domains. This is linked to the role played by convolutional layers at extracting image features and reducing the data complexity. In this case, the number of synaptic weights can be reduced in comparison to conventional MLPs.}, language = {en} } @misc{SoltaniZarrinZahariMahadevaiahetal., author = {Soltani Zarrin, Pouya and Zahari, Finn and Mahadevaiah, Mamathamba Kalishettyhalli and P{\´e}rez, Eduardo and Kohlstedt, Hermann and Wenger, Christian}, title = {Neuromorphic on‑chip recognition of saliva samples of COPD and healthy controls using memristive devices}, series = {Scientific Reports}, volume = {10}, journal = {Scientific Reports}, issn = {2045-2322}, doi = {10.1038/s41598-020-76823-7}, abstract = {Chronic Obstructive Pulmonary Disease (COPD) is a life-threatening lung disease, affecting millions of people worldwide. Implementation of Machine Learning (ML) techniques is crucial for the effective management of COPD in home-care environments. However, shortcomings of cloud-based ML tools in terms of data safety and energy efficiency limit their integration with low-power medical devices. To address this, energy efficient neuromorphic platforms can be used for the hardware-based implementation of ML methods. Therefore, a memristive neuromorphic platform is presented in this paper for the on-chip recognition of saliva samples of COPD patients and healthy controls. The results of its performance evaluations showed that the digital neuromorphic chip is capable of recognizing unseen COPD samples with accuracy and sensitivity values of 89\% and 86\%, respectively. Integration of this technology into personalized healthcare devices will enable the better management of chronic diseases such as COPD.}, language = {en} } @misc{ZahariPerezMahadevaiahetal., author = {Zahari, Finn and P{\´e}rez, Eduardo and Mahadevaiah, Mamathamba Kalishettyhalli and Kohlstedt, Hermann and Wenger, Christian and Ziegler, Martin}, title = {Analogue pattern recognition with stochastic switching binary CMOS‑integrated memristive devices}, series = {Scientific Reports}, volume = {10}, journal = {Scientific Reports}, issn = {2045-2322}, doi = {10.1038/s41598-020-71334-x}, pages = {15}, abstract = {Biological neural networks outperform todays computer technology in terms of power consumption and computing speed when associative tasks, like pattern recognition, are to be solved. The analogue and massive parallel in-memory computing in biology differs strongly with conventional transistor electronics using the von Neumann architecture. Therefore, novel bio-inspired computing architectures are recently highly investigated in the area of neuromorphic computing. Here, memristive devices, which serve as non-volatile resistive memory, are used to emulate the plastic behaviour of biological synapses. In particular, CMOS integrated resistive random access memory (RRAM) devices are promising candidates to extend conventional CMOS technology in neuromorphic systems. However, dealing with the inherent stochasticity of the resistive switching effect can be challenging for network performance. In this work, the probabilistic switching is exploited to emulate stochastic plasticity with fully CMOS integrated binary RRAM devices. Two different RRAM technologies with different device variabilities are investigated in detail and their use in a stochastic artificial neural network (StochANN) to solve the MINST pattern recognition task is examined. A mixed-signal implementation with hardware synapses and software neurons as well as numerical simulations show the proposed concept of stochastic computing is able to handle analogue data with binary memory cells.}, language = {en} } @misc{MiloAnzaloneZambellietal., author = {Milo, Valerio and Anzalone, Francesco and Zambelli, Cristian and P{\´e}rez, Eduardo and Mahadevaiah, Mamathamba Kalishettyhalli and Ossorio, {\´O}scar G. and Olivo, Piero and Wenger, Christian and Ielmini, Daniele}, title = {Optimized programming algorithms for multilevel RRAM in hardware neural networks}, series = {IEEE International Reliability Physics Symposium (IRPS), 2021}, journal = {IEEE International Reliability Physics Symposium (IRPS), 2021}, isbn = {978-1-7281-6894-4}, issn = {1938-1891}, doi = {10.1109/IRPS46558.2021.9405119}, abstract = {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.}, language = {en} } @misc{RomeroZalizPerezJimenezMolinosetal., author = {Romero-Zaliz, Roc{\´i}o and P{\´e}rez, Eduardo and Jimenez-Molinos, Francisco and Wenger, Christian and Roldan, Juan Bautista}, title = {Influence of variability on the performance of HfO2 memristor-based convolutional neural networks}, series = {Solid State Electronics}, volume = {185}, journal = {Solid State Electronics}, issn = {0038-1101}, doi = {10.1016/j.sse.2021.108064}, pages = {5}, abstract = {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.}, language = {en} } @misc{BaroniZambelliOlivoetal., author = {Baroni, Andrea and Zambelli, Cristian and Olivo, Piero and P{\´e}rez, Eduardo and Wenger, Christian and Ielmini, Daniele}, title = {Tackling the Low Conductance State Drift through Incremental Reset and Verify in RRAM Arrays}, series = {2021 IEEE International Integrated Reliability Workshop (IIRW), South Lake Tahoe, CA, USA, 10 December 2021}, journal = {2021 IEEE International Integrated Reliability Workshop (IIRW), South Lake Tahoe, CA, USA, 10 December 2021}, publisher = {Institute of Electrical and Electronics Engineers (IEEE)}, isbn = {978-1-6654-1794-5}, issn = {2374-8036}, doi = {10.1109/IIRW53245.2021.9635613}, pages = {5}, abstract = {Resistive switching memory (RRAM) is a promising technology for highly efficient computing scenarios. RRAM arrays enabled the acceleration of neural networks for artificial intelligence and the creation of In-Memory Computing circuits. However, the arrays are affected by several issues materializing in conductance variations that might cause severe performance degradation in those applications. Among those, one is related to the drift of the low conductance states appearing immediately at the end of program and verify algorithms that are fundamental for an accurate Multi-level conductance operation. In this work, we tackle the issue by developing an Incremental Reset and Verify technique showing enhanced variability and reliability features compared with a traditional refresh-based approach.}, language = {en} } @misc{RomeroZalizCantudoPerezetal., author = {Romero-Zaliz, Rocio and Cantudo, Antonio and P{\´e}rez, Eduardo and Jimenez-Molinos, Francisco and Wenger, Christian and Roldan, Juan Bautista}, title = {An Analysis on the Architecture and the Size of Quantized Hardware Neural Networks Based on Memristors}, series = {Electronics (MDPI)}, volume = {10}, journal = {Electronics (MDPI)}, number = {24}, issn = {2079-9292}, doi = {10.3390/electronics10243141}, abstract = {We have performed different simulation experiments in relation to hardware neural networks (NN) to analyze the role of the number of synapses for different NN architectures in the network accuracy, considering different datasets. A technology that stands upon 4-kbit 1T1R ReRAM arrays, where resistive switching devices based on HfO2 dielectrics are employed, is taken as a reference. In our study, fully dense (FdNN) and convolutional neural networks (CNN) were considered, where the NN size in terms of the number of synapses and of hidden layer neurons were varied. CNNs work better when the number of synapses to be used is limited. If quantized synaptic weights are included, we observed thatNNaccuracy decreases significantly as the number of synapses is reduced; in this respect, a trade-off between the number of synapses and the NN accuracy has to be achieved. Consequently, the CNN architecture must be carefully designed; in particular, it was noticed that different datasets need specific architectures according to their complexity to achieve good results. It was shown that due to the number of variables that can be changed in the optimization of a NN hardware implementation, a specific solution has to be worked in each case in terms of synaptic weight levels, NN architecture, etc.}, language = {en} } @misc{MannocciBaroniMelacarneetal., author = {Mannocci, Piergiulio and Baroni, Andrea and Melacarne, Enrico and Zambelli, Cristian and Olivo, Piero and P{\´e}rez, Eduardo and Wenger, Christian and Ielmini, Daniele}, title = {In-Memory Principal Component Analysis by Crosspoint Array of Rresistive Switching Memory}, series = {IEEE Nanotechnology Magazine}, volume = {16}, journal = {IEEE Nanotechnology Magazine}, number = {2}, issn = {1932-4510}, doi = {10.1109/MNANO.2022.3141515}, pages = {4 -- 13}, abstract = {In Memory Computing (IMC) is one of the most promising candidates for data-intensive computing accelerators of machine learning (ML). A key ML algorithm for dimensionality reduction and classification is principal component analysis (PCA), which heavily relies on matrixvector multiplications (MVM) for which classic von Neumann architectures are not optimized. Here, we provide the experimental demonstration of a new IMCbased PCA algorithm based on power iteration and deflation executed in a 4-kbit array of resistive switching random-access memory (RRAM). The classification accuracy of the Wisconsin Breast Cancer data set reaches 95.43\%, close to floatingpoint implementation. Our simulations indicate a 250× improvement in energy efficiency compared to commercial GPUs, thus supporting IMC for energy-efficient ML in modern data-intensive computing.}, language = {en} } @misc{FritscherKnoedtelMallahetal., author = {Fritscher, Markus and Kn{\"o}dtel, Johannes and Mallah, Maen and Pechmann, Stefan and Perez-Bosch Quesada, Emilio and Rizzi, Tommaso and Wenger, Christian and Reichenbach, Marc}, title = {Mitigating the Effects of RRAM Process Variation on the Accuracy of Artifical Neural Networks}, series = {Embedded Computer Systems: Architectures, Modeling, and Simulation. SAMOS 2021. Lecture Notes in Computer Science}, journal = {Embedded Computer Systems: Architectures, Modeling, and Simulation. SAMOS 2021. Lecture Notes in Computer Science}, publisher = {Springer}, isbn = {978-3-031-04579-0}, issn = {0302-9743}, doi = {10.1007/978-3-031-04580-6_27}, pages = {401 -- 417}, abstract = {Weight storage is a key challenge in the efficient implementation of artificial neural networks. Novel memory technologies such as RRAM are able to greatly improve density and introduce non-volatility and multibit capabilities to this component of ANN accelerators. The usage of RRAM in this domain comes with downsides, mainly caused by cycle-to-cycle and device-to-device variability leading to erroneous readouts, greatly affecting digital systems. ANNs have the ability to compensate for this by their inherent redundancy and usually exhibit a gradual deterioration in the accuracy of the task at hand. This means, that slight error rates can be acceptable for weight storage in an ANN accelerator. In this work we link device-to-device variability to the accuracy of an ANN for such an accelerator. From this study, we can estimate how strongly a certain net is affected by a certain device parameter variability. This methodology is then used to present three mitigation strategies and to evaluate how they affect the reaction of the network to variability: a) Dropout Layers b) Fault-Aware Training c) Redundancy. These mitigations are then evaluated by their ability to improve accuracy and to lower hardware overhead by providing data for a real-word example. We improved this network's resilience in such a way that it could tolerate double the variation in one of the device parameters (standard deviation of the oxide thickness can be 0.4 nm instead of 0.2 nm while maintaining sufficient accuracy.)}, language = {en} } @misc{GlukhovMiloBaronietal., author = {Glukhov, Artem and Milo, Valerio and Baroni, Andrea and Lepri, Nicola and Zambelli, Cristian and Olivo, Piero and P{\´e}rez, Eduardo and Wenger, Christian and Ielmini, Daniele}, title = {Statistical model of program/verify algorithms in resistive-switching memories for in-memory neural network accelerators}, series = {2022 IEEE International Reliability Physics Symposium (IRPS)}, journal = {2022 IEEE International Reliability Physics Symposium (IRPS)}, publisher = {Institute of Electrical and Electronics Engineers (IEEE)}, isbn = {978-1-6654-7950-9}, issn = {2473-2001}, doi = {10.1109/IRPS48227.2022.9764497}, pages = {3C.3-1 -- 3C.3-7}, abstract = {Resistive-switching random access memory (RRAM) is a promising technology for in-memory computing (IMC) to accelerate training and inference of deep neural networks (DNNs). This work presents the first physics-based statistical model describing (i) multilevel RRAM device program/verify (PV) algorithms by controlled set transition, (ii) the stochastic cycle-to-cycle (C2C) and device-to-device (D2D) variations within the array, and (iii) the impact of such imprecisions on the accuracy of DNN accelerators. The model can handle the full chain from RRAM materials/device parameters to the DNN performance, thus providing a valuable tool for device/circuit codesign of hardware DNN accelerators.}, language = {en} } @misc{BogunPerezBoschQuesadaPerezetal., author = {Bogun, Nicolas and Perez-Bosch Quesada, Emilio and P{\´e}rez, Eduardo and Wenger, Christian and Kloes, Alexander and Schwarz, Mike}, title = {Analytical Calculation of Inference in Memristor-based Stochastic Artificial Neural Networks}, series = {29th International Conference on Mixed Design of Integrated Circuits and System (MIXDES), 23-24 June 2022 , Wrocław, Poland}, journal = {29th International Conference on Mixed Design of Integrated Circuits and System (MIXDES), 23-24 June 2022 , Wrocław, Poland}, isbn = {978-83-63578-22-0}, doi = {10.23919/MIXDES55591.2022.9838321}, pages = {83 -- 88}, abstract = {The impact of artificial intelligence on human life has increased significantly in recent years. However, as the complexity of problems rose aswell, increasing system features for such amount of data computation became troublesome due to the von Neumann's computer architecture. Neuromorphic computing aims to solve this problem by mimicking the parallel computation of a human brain. For this approach, memristive devices are used to emulate the synapses of a human brain. Yet, common simulations of hardware based networks require time consuming Monte-Carlo simulations to take into account the stochastic switching of memristive devices. This work presents an alternative concept making use of the convolution of the probability distribution functions (PDF) of memristor currents by its equivalent multiplication in Fourier domain. An artificial neural network is accordingly implemented to perform the inference stage with handwritten digits.}, language = {en} } @misc{BaroniGlukhovPerezetal., author = {Baroni, Andrea and Glukhov, Artem and P{\´e}rez, Eduardo and Wenger, Christian and Ielmini, Daniele and Olivo, Piero and Zambelli, Cristian}, title = {Low Conductance State Drift Characterization and Mitigation in Resistive Switching Memories (RRAM) for Artificial Neural Networks}, series = {IEEE Transactions on Device and Materials Reliability}, volume = {22}, journal = {IEEE Transactions on Device and Materials Reliability}, number = {3}, issn = {1530-4388}, doi = {10.1109/TDMR.2022.3182133}, pages = {340 -- 347}, abstract = {The crossbar structure of Resistive-switching random access memory (RRAM) arrays enabled the In-Memory Computing circuits paradigm, since they imply the native acceleration of a crucial operations in this scenario, namely the Matrix-Vector-Multiplication (MVM). However, RRAM arrays are affected by several issues materializing in conductance variations that might cause severe performance degradation. A critical one is related to the drift of the low conductance states appearing immediately at the end of program and verify algorithms that are mandatory for an accurate multi-level conductance operation. In this work, we analyze the benefits of a new programming algorithm that embodies Set and Reset switching operations to achieve better conductance control and lower variability. Data retention analysis performed with different temperatures for 168 hours evidence its superior performance with respect to standard programming approach. Finally, we explored the benefits of using our methodology at a higher abstraction level, through the simulation of an Artificial Neural Network for image recognition task (MNIST dataset). The accuracy achieved shows higher performance stability over temperature and time.}, language = {en} } @misc{GlukhovLepriMiloetal., author = {Glukhov, Artem and Lepri, Nicola and Milo, Valerio and Baroni, Andrea and Zambelli, Cristian and Olivo, Piero and P{\´e}rez, Eduardo and Wenger, Christian and Ielmini, Daniele}, title = {End-to-end modeling of variability-aware neural networks based on resistive-switching memory arrays}, series = {Proc. 30th IFIP/IEEE International Conference on Very Large Scale Integration (VLSI-SoC 2022)}, journal = {Proc. 30th IFIP/IEEE International Conference on Very Large Scale Integration (VLSI-SoC 2022)}, doi = {10.1109/VLSI-SoC54400.2022.9939653}, pages = {1 -- 5}, abstract = {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.}, language = {en} }