arXiv subjects

Adnan Mehonic

Publications and source records attributed to Adnan Mehonic.

9 recordsLinked to original sources

arXiv2026

Low-power analogue neural networks with trainable nonlinear connections for continuous control

Physical neural networks promise low-power machine learning by computing directly with analogue device physics, but most architectures force nonlinear device responses to act as scalar weights. Inspired by Kolmogorov-Arnold networks, we place trainable nonlinear functions on the connections, making each physical connection a learnable computational element. Realising these functions as analogue band-pass filters on field-programmable analogue arrays, we find that the benefit is task-dependent and follows from the smoothness of the physical basis: the networks represent smooth, continuously valued targets, including robotic kinematics, continuous control, and photovoltaic maximum-power-point tracking, with far fewer nodes and connections than multilayer perceptrons, but offer no parameter-efficiency advantage on classification-like decision boundaries. Trained networks transfer to hardware across approximately 35,000 connections with quantified fidelity, and a dedicated CMOS implementation is projected to operate at approximately 30 microwatts. A memristive realisation reproduces the same behaviour in simulation, indicating that the advantage comes from placing trainable nonlinearity on connections, rather than from a particular device.

cs.LG↗

arXiv2024

Nonideality-aware training makes memristive networks more robust to adversarial attacks

Neural networks are now deployed in a wide number of areas from object classification to natural language systems. Implementations using analog devices like memristors promise better power efficiency, potentially bringing these applications to a greater number of environments. However, such systems suffer from more frequent device faults and overall, their exposure to adversarial attacks has not been studied extensively. In this work, we investigate how nonideality-aware training - a common technique to deal with physical nonidealities - affects adversarial robustness. We find that adversarial robustness is significantly improved, even with limited knowledge of what nonidealities will be encountered during test time.

cs.ET↗

arXiv2024

Roadmap to Neuromorphic Computing with Emerging Technologies

The roadmap is organized into several thematic sections, outlining current computing challenges, discussing the neuromorphic computing approach, analyzing mature and currently utilized technologies, providing an overview of emerging technologies, addressing material challenges, exploring novel computing concepts, and finally examining the maturity level of emerging technologies while determining the next essential steps for their advancement.

eess.SP↗

arXiv2023

Emerging Nonvolatile Memories for Machine Learning

Digital computers have been getting exponentially faster for decades, but huge challenges exist today. Transistor scaling, described by Moore's law, has been slowing down over the last few years, ending the era of fully predictable performance improvements. Furthermore, the data-centric computing demands fueled by machine learning applications are rapidly growing, and current computing systems -- even with the historical rate of improvements driven by Moore's law -- cannot keep up with these enormous computational demands. Some are turning to analogue in-memory computing as a solution, where specialised systems operating on physical principles accelerate specific tasks. We explore how emerging nonvolatile memories can be used to implement such systems tailored for machine learning. In particular, we discuss how memristive crossbar arrays can accelerate key linear algebra operations used in neural networks, what technological challenges remain, and how they can be overcome.

cs.ET↗

arXiv2022

Memristive, Spintronic, and 2D-Materials-Based Devices to Improve and Complement Computing Hardware

In a data-driven economy, virtually all industries benefit from advances in information technology -- powerful computing systems are critically important for rapid technological progress. However, this progress might be at risk of slowing down if we do not address the discrepancy between our current computing power demands and what the existing technologies can offer. Key limitations to improving energy efficiency are the excessive growth of data transfer costs associated with the von Neumann architecture and the fundamental limits of complementary metal-oxide-semiconductor (CMOS) technologies, such as transistors. In this perspective article, we discuss three technologies that will likely play an essential role in future computing systems: memristive electronics, spintronics, and electronics based on 2D materials. We present how these may transform conventional digital computers and contribute to the adoption of new paradigms, like neuromorphic computing.

cs.ET↗

arXiv2022

Nonideality-Aware Training for Accurate and Robust Low-Power Memristive Neural Networks

Recent years have seen a rapid rise of artificial neural networks being employed in a number of cognitive tasks. The ever-increasing computing requirements of these structures have contributed to a desire for novel technologies and paradigms, including memristor-based hardware accelerators. Solutions based on memristive crossbars and analog data processing promise to improve the overall energy efficiency. However, memristor nonidealities can lead to the degradation of neural network accuracy, while the attempts to mitigate these negative effects often introduce design trade-offs, such as those between power and reliability. In this work, we design nonideality-aware training of memristor-based neural networks capable of dealing with the most common device nonidealities. We demonstrate the feasibility of using high-resistance devices that exhibit high $I$-$V$ nonlinearity -- by analyzing experimental data and employing nonideality-aware training, we estimate that the energy efficiency of memristive vector-matrix multipliers is improved by three orders of magnitude ($0.715\ \mathrm{TOPs}^{-1}\mathrm{W}^{-1}$ to $381\ \mathrm{TOPs}^{-1}\mathrm{W}^{-1}$) while maintaining similar accuracy. We show that associating the parameters of neural networks with individual memristors allows to bias these devices towards less conductive states through regularization of the corresponding optimization problem, while modifying the validation procedure leads to more reliable estimates of performance. We demonstrate the universality and robustness of our approach when dealing with a wide range of nonidealities.

cs.ET↗

arXiv2021

Brain-inspired computing: We need a master plan

New computing technologies inspired by the brain promise fundamentally different ways to process information with extreme energy efficiency and the ability to handle the avalanche of unstructured and noisy data that we are generating at an ever-increasing rate. To realise this promise requires a brave and coordinated plan to bring together disparate research communities and to provide them with the funding, focus and support needed. We have done this in the past with digital technologies; we are in the process of doing it with quantum technologies; can we now do it for brain-inspired computing?

cs.ET↗

arXiv2020

Nanoscale plasticity and neuromorphic dynamics in silicon suboxide RRAM

Resistive random-access memories, also known as memristors, whose resistance can be modulated by the electrically driven formation and disruption of conductive filaments within an insulator, are promising candidates for neuromorphic applications due to their scalability, low-power operation and diverse functional behaviours. However, understanding the dynamics of individual filaments, and the surrounding material, is challenging, owing to the typically very large cross-sectional areas of test devices relative to the nanometre scale of individual filaments. In the present work, conductive atomic force microscopy is used to study the evolution of conductivity at the nanoscale in a fully CMOS-compatible silicon suboxide thin film. Distinct filamentary plasticity and background conductivity enhancement are reported, suggesting that device behaviour might be best described by composite core (filament) and shell (background conductivity) dynamics. Furthermore, constant current measurements demonstrate an interplay between filament formation and rupture, resulting in current-controlled voltage spiking in nanoscale regions, with an estimated optimal energy consumption of 25 attojoules per spike. This is very promising for extremely low-power neuromorphic computation and suggests that the dynamic behaviour observed in larger devices should persist and improve as dimensions are scaled down.

cond-mat.mtrl-sci↗

arXiv2020

Memristors -- from In-memory computing, Deep Learning Acceleration, Spiking Neural Networks, to the Future of Neuromorphic and Bio-inspired Computing

Machine learning, particularly in the form of deep learning, has driven most of the recent fundamental developments in artificial intelligence. Deep learning is based on computational models that are, to a certain extent, bio-inspired, as they rely on networks of connected simple computing units operating in parallel. Deep learning has been successfully applied in areas such as object/pattern recognition, speech and natural language processing, self-driving vehicles, intelligent self-diagnostics tools, autonomous robots, knowledgeable personal assistants, and monitoring. These successes have been mostly supported by three factors: availability of vast amounts of data, continuous growth in computing power, and algorithmic innovations. The approaching demise of Moore's law, and the consequent expected modest improvements in computing power that can be achieved by scaling, raise the question of whether the described progress will be slowed or halted due to hardware limitations. This paper reviews the case for a novel beyond CMOS hardware technology, memristors, as a potential solution for the implementation of power-efficient in-memory computing, deep learning accelerators, and spiking neural networks. Central themes are the reliance on non-von-Neumann computing architectures and the need for developing tailored learning and inference algorithms. To argue that lessons from biology can be useful in providing directions for further progress in artificial intelligence, we briefly discuss an example based reservoir computing. We conclude the review by speculating on the big picture view of future neuromorphic and brain-inspired computing systems.

cs.ET↗