Plenary Conférence 1

In-Memory Computing: A Path to Energy-Efficienty and Trustworthy Embedded AI 

Dr. Damien QUERLIOZ

Centre de Nanosciences et de Nanotechnologies, Univ. Paris-Saclay, France

Damien Querlioz is a CNRS Research Director at the Centre de Nanosciences et de Nanotechnologies of Université Paris-Saclay and CNRS. His research focuses on novel usages of emerging non-volatile memory and other nanodevices, in particular relying on inspirations from biology and machine learning. He received his predoctoral education at Ecole Normale Supérieure, Paris and his PhD from Université Paris-Sud in 2009. Before his appointment at CNRS, he was a Postdoctoral Scholar at Stanford University and at the Commissariat à l'Energie Atomique. Damien Querlioz is the coordinator of the interdisciplinary INTEGNANO research group, with colleagues working on all aspects of nanodevice physics and technology, from materials to systems. He has co-authored one book, nine book chapters, more than 150 journal articles, and conference proceedings, and given more than 80 invited talks at national and international workshops and conferences. In 2016, he was the recipient of an ERC Starting Grant to develop the concept of natively intelligent memory. In 2017, he received the CNRS Bronze medal. He has also been a co-recipient of the 2017 IEEE Guillemin-Cauer Best Paper Award and of the 2018 IEEE Biomedical Circuits and Systems Best Paper Award.

Abstract

 Artificial intelligence (AI) has immense potential for edge applications, but its energy demands are a critical barrier to widespread adoption in fields such as medical implants and advanced brain-machine interfaces. Traditional computing systems face substantial inefficiencies due to the high energy costs of memory access. In contrast, the human brain achieves remarkable energy efficiency by performing computation directly within memory structures. This talk explores innovative approaches to embedded AI through in-memory and near-memory computing, focusing on integrating logic and memory to drastically reduce energy consumption. We highlight the development and application of emerging non-volatile memory technologies like memristors, magnetic memory, and phase-change memory. These technologies emulate the brain's energy-efficient architecture and have recently reached maturity, enabling the demonstration of fully functional in-memory computing systems, which we will showcase throughout this talk. These advanced memory technologies facilitate the creation of digital, low-precision neural networks, offering robust, low-power solutions for AI inference. They also enable analog in-memory computing, which naturally performs neural network operations through fundamental electrical laws. Despite their potential, these technologies come with significant challenges due to their variability. We demonstrate how Bayesian techniques can not only tolerate these imperfections but sometimes even leverage them. The presentation will also cover recent advances in local learning algorithms, such as Equilibrium Propagation, which promise efficient on-chip learning capabilities using in-memory computing. Finally, we will discuss the current challenges and future directions for incorporating such technologies at the architectural level.

Plenary Conférence 2

Introduction to Quantum Computing

Pr. Mourad TELMINI

Univ. Tunis El Manar  (Tunisia)

Mourad Telmini is Full Professor of physics at the Faculty of Sciences in Tunis (University of Tunis El Manar) and director of the Laboratory for Atomic and Molecular Spectroscopy & Applications (LSAMA). He holds a PhD in Atomic Physics from the University of Paris-Sud. He teaches quantum mechanics, atomic physics and quantum information in undergraduate and master’s degrees. His research activities concern atomic physics (cold atoms and Bose-Einstein condensates in optical networks) and molecular (Rydberg states, halfium model, coherent control by ultra-short laser pulses). More recently, he has started research in quantum computing (variational calculus with quantum algorithms). He has published dozens of articles in international journals and has participated as a guest speaker at several international conferences. He is visiting professor at several universities in France, Germany, Brazil, South Africa, etc. He has held the positions of Director General of the National Centre for Nuclear Sciences and Technologies (CNSTN), and Vice President of the African Commission on Nuclear Energy (AFCONE). He represents Tunisia in the International Union of Pure and Applied Physics (IUPAP) on behalf of the Tunisian Society of Physics (STP), of which he is Vice-President, and currently chairs the Tunisian Association of Metrology (ATMET). He is also editor of the magazine African Physics Newsletters and co-head of the Tunisian quantum network QUANTUN, representing Tunisia in the international World Quantum Day network

Abstract

Quantum Computing is one of the most trending and promising chapters of all quantum technologies. The basic idea behind its development is the possibility to rely on the quantum properties of matter at the microscopic scale, mainly quantum superposition and quantum entanglement, in order to build up computing hardware (quantum circuits) and software (quantum algorithms) that can handle complex problems which are out of reach of conventional computing resources, within a reasonable amount of time.

Quantum superposition has been demonstrated a long time ago and routinely used in several applications such as atomic clocks and interferometry, using, inter alia, the basic Rabi oscillations phenomenon. However, quantum entanglement was by far a more elusive phenomenon that has required significantly more effort from the physics community to provide fully accepted evidence of it. This field was triggered by the groundbreaking experimental work of A. Aspect in the 1980's, which demonstrated the violation of Bell's inequalities and proved the existence of quantum entanglement. Further research conducted by A. Zeilinger and others was able to implement quantum teleportation, based on quantum entanglement, opening up the possibility of effectively considering technological applications of these quantum properties.

In this talk, I will present an overview of Quantum computing, including state-of-the-art developments. Then, I will demonstrate the power of quantum algorithms in the case of quantum cryptography, based on the Quantum Key Distribution (QKD) protocol.

Furthermore, In will say a few words on the research that we are conducting in the laboratory of Atomic and Molecular Spectroscopy & Applications (LSAMA) in the field of hybrid algorithms, like the Variational Quantum Eigensolver (VQE), and their use in the electronic structure calculations in atomic and molecular physics

Plenary Conférence 3

Advanced characterization of interface traps in electronic & photonic III-V devices

Pr. Paul HURLEY

Tyndall National Institute (Ireland)

Prof. Paul Hurley is currently the Head of the Nanoelectronic Materials and Devices Group at the Tyndall National Institute and Professor at University College Cork. Paul leads a research group of around 25 PhD students, post-doctoral researchers, visiting students and Tyndall Research staff who perform research into alternative semiconductor materials and device structures aimed at improving the energy efficiency in the next generation of logic devices. Paul’s group investigates synaptic transistors using 2D materials, 2D & Oxide Semiconductors for electronic device applications, resistive switching in single/few layer 2 D materials, Ions in oxides as a new class of memory to realise dedicated neuromorphic hardware for the development of next generation integrated circuits (ICs). Paul is currently leading a four-year programme of research funded by Meta to enhance the efficiency of micro LEDs for display technologies and AR applications.

Paul received an Intel Outstanding Researcher award for his work on electrically active defects in high-k/III-V metal-oxide-semiconductor (MOS) system in 2012.

Abstract

The ability to understand and control the properties of semiconductor/insulator interfaces has been essential for the development of transistor and memory technology in the last 60 years. The primary method used to analyse the properties of semiconductor/oxide interfaces in metal-oxide-semiconductor (MOS) structures is through the measurement and analysis of the small signal ac impedance of the MOS system. The measured impedance (Z) and phase angle (q) are converted to an equivalent capacitance (C) and conductance (G). Deviations of the measured C and G values from the ideal values as a function of the applied voltage and the ac signal frequency and used to characterise electrically active defects in the MOS system. In the case of the silicon MOS system the impedance spectroscopy is well established, and the dominant electrically active defects are located at the silicon/oxide interface. In the case of MOS structures formed on III-V compound semiconductors, a range of scientific and technological questions remain. This presentation will cover the application of impedance spectroscopy to III-V MOS structures relevant to electronic (InGaAs) and opto-electronic (InGaP) applications. The multi-frequency capacitance-voltage (CV) and conductance-voltage (GV) of the InGaAs MOS system have been analysed by many groups.  The presentation will demonstrate how a complete description of the CV and GV characteristics of the InGaAs MOS system, requires that the projection of all defects to the InGaAs/oxide interface must be abandoned, and multi-phonon non-local tunnelling to defects in the oxide is required to accurately fit the experimental data. For InGaP structures in microLEDs, the oxide/InGaP interface at the edge of the mesa structure plays a dominant role in non-radiative electron-hole pair recombination.  Finally, the presentation will cover initial studies of impedance spectroscopy applied to the InGaP MOS system. 

Plenary Conférence 4

State of the art on WBG and UWBG

Power Semiconductor Devices

 Dr. Pierre BROSSELARD

Ampere Laboratory

INSA Lyon (France)

Pierre Brosselard was born in Roanne, France, in 1977. He received the B.S. degree from the University of Lyon, Lyon, France, in 2001, and the Ph.D. degree from Institut National des Sciences Apliquées de Lyon, Lyon, in 2004. From 2001 to 2005, he was with the Centre de Génie Eléctrique de Lyon, Equipe de Composants de Puissance et Applications, Lyon, where he worked on SiC power devices. From 2005 to 2008, he was with the Power Electronics Group, Instituto de Microelectrònica de Barcelona-Centro Nacional de Microelectrònica, Consejo Superior de Investigaciones Científica, Bellaterra, Spain. From 2008 to 2014, he has been an Assistant Professor with the Institut National des Sciences Appliquées de Lyon, Villeurbanne, France and AMPERE-lab. From 2014 to 2022, he has co-founded CALY Technologies a spin-off of AMPERE-lab. Since 2022, he joined AMPERE-lab. He has authored and coauthored more than 90 research papers in journals and conference proceedings. His current research activity deals with the modeling, design, and characterization of power-semiconductor devices.

Abstract

Wide Bang Gap (WBG) and Ultra-Wide Band Gap (UWBG) semiconductor devices are very promising for power electronic applications. Actually, Silicon Carbide (SiC) and Gallium Nitride (GaN) power devices are available on market. Infineon, STMicroelectronics, Wolfspeed, ROHM and others sell devices. The maximum breakdown voltage is 650V for GaN switches and 3300V for SiC ones. In parallel, research works on reliability and evolution of the design aspects are in progress for that Transistors and Diodes. For higher blocking voltage, some prototypes are available but more design optimizations are doing. The WBG semiconductor technology is optimized but for thick material, solutions must be developed to reduce the cost. For UWBG, the quality of the material and the technology are not so mature. Only proof of concept is available for Diamond, Ga2O3 and AlN devices. Regarding the material aspect, some companies commercialized substrates. In this paper, we will discuss about these different aspects.

Plenary Conférence 5

Morocco's green hydrogen opportunity

 Dr. Hicham BOUZEKRI

African Technical Advisors (Morocco)

Abstract

Morocco is on the verge of becoming a net energy exporter for the first time since its independence thanks to the promise of the national Green Hydrogen Offer initiative. Along with this great opportunity comes significant challenges to translate this energy export dynamic to permanent job creation opportunities and transform the country's industrial landscape. Through a detailed analysis of Morocco's green hydrogen offer and taking into account an international benchmark, this talk will attempt to highlight strengths, potential economic impacts and threats of this promising phase of the euro-mediterranean energy transition.

Plenary Conférence 6

Future Computing Chips: The Need for a New Mindset to Sustain the World in the Age of AI

Pr. Saïd HAMDIOUI 

Computer Engineering Laboratory

TU Delft (Netherlands)

Said Hamdioui (http://www.ce.ewi.tudelft.nl/hamdioui) is Chair Professor of Dependable and Emerging Computer Technologies and Head of the Computer Engineering Laboratory at Delft University of Technology (TU Delft), Netherlands.His research spans emerging technologies and computing paradigms, such as in-memory and brain-inspired computing, as well as hardware dependability, including testability, reliability, and security. Before joining academia, he gained extensive industry experience, working with the Microprocessor Products Group at Intel Corporation (California, USA), the IP and Yield Group at Philips Semiconductors R&D (Crolles, France), and the DSP Design Group at Philips/NXP Semiconductors (Nijmegen, Netherlands). Hamdioui holds four patents, authored one book, contributed to two others, and published over 330 peer-reviewed papers. He has provided consulting and training for leading semiconductor companies, including Intel, NXP, STMicroelectronics, Renesas, and Huawei, with a focus on digital integrated circuit testing and design-for-test. Additionally, he has collaborated extensively with top academic and industrial partners such as IBM, IMEC, Cadence, the European Space Agency, ETH Zurich, and Politecnico di Torino. His contributions have significantly advanced the fields of chip design, semiconductors, and Electronic Design Automation (EDA).

Said Hamdioui is an active contributor to the international research community, serving on organizing and technical program committees for leading conferences. He has delivered numerous keynote speeches, distinguished lectures, and invited tutorials at major forums, conferences, and semiconductor companies. His editorial roles include Associate Editor for IEEE Transactions on VLSI Systems (2015–2018), and editorial contributions to the Microelectronics Reliability Journal (2019–2020) and JETTA (2011–2019). He currently serves on the editorial boards of the ACM Journal on Emerging Technologies in Computing Systems (2020–present) and IEEE Design & Test (2013–present).

Said Hamdioui has received numerous awards, including the EDAA Outstanding Dissertation Award (2001) for his pioneering work on memory testing, the European Commission Innovation Award (2020) for the H2020 MNEMOSENE project on energy-efficient computing, and HiPEAC Technology Transfer Awards (2015, 2022). He was also recognized as the Best Tech Idea of the Netherlands (2021), received over 20 Best Paper Awards and nominations at prestigious conferences like DATE, ITC, and ETS, and was named Teacher of the Year (2017) at TU Delft. He is an IEEE Circuits and Systems Society Distinguished Lecturer (2021–2022) and a key member of the Cadence Academic Network on Dependability and Design-for-Testability. Ranked among the World’s Top 2% Scientists (Stanford & Elsevier 2024), he is a Fellow of the Netherlands Academy of Engineering and a Senior Member of IEEE. Hamdioui also serves on the AENEAS/ENIAC Scientific Committee and advisory boards for Khalifa University (UAE) and the Graduate School for Intelligent Methods for Test and Reliability, a collaboration between Advantest and the University of Stuttgart.

Abstract 

The pervasive role of electronic systems in our daily lives is undeniable. From smartphones and computers to televisions and coffee machines, modern life relies heavily on electronics and the internet. Without these technologies, business operations would grind to a halt, educational quality would decline, and daily life would regress to pre-industrial standards. However, this dependence comes at a significant and growing cost, particularly with the rise of power-hungry applications like Artificial Intelligence (AI). For instance, energy forecasts predict that the electricity demand for ICT systems will exceed 20% of global electricity consumption by 2030, driven by data centers, smartphones, computers, and networks. This highlights the urgent need to design and use these systems in an energy-efficient manner, not only to reduce ICT’s environmental footprint and ensure sustainability but also to enable emerging energy-constrained applications, such as edge AI.

This talk will explore the current state of chip technology and computer hardware architectures, which form the backbone of ICT systems, and examine their limitations in achieving the energy efficiency required for a sustainable future. It will address both device-level and architectural challenges, emphasizing the need for innovative solutions. A particular focus will be on dedicated hardware architectures for AI, such as Computation-In-Memory (CIM) architectures that leverage memristor devices and draw inspiration from the human brain. The immense potential of CIM, capable of delivering over 100x improvements in energy efficiency, will be illustrated through real-world case studies and experimental data from chip prototypes. Additionally, key aspects related to the design, testing, and reliability of brain-inspired CIM architectures will be discussed, along with future challenges in chip technology and computer hardware design.