top of page



Pivoting for the Future: How Melexis Is Adapting and Innovating for the Future

Tuesday, September 10



Innovation in sensors stands at a crossroads. On one hand, the rise of artificial intelligence is driven by advanced digital electronics chips implemented in cutting-edge CMOS technologies which are at the center of geopolitical tensions. On the other hand, these intelligent systems lack what we might call "physical intelligence". While a robot can articulate speech in all languages, it still faces challenges when it comes to opening a door – a phenomenon known as Moravec's paradox. The ability to make sense of the environment and autonomously manipulate objects relies on smart sensors in legacy technologies, driven by both technical and business reasons. To bridge the widening gap between artificial and physical intelligence, there is a need for sensor innovation across various dimensions : sensing modalities, sensor technology nodes, and sensor architecture. Breakthroughs are needed to enable sensorimotor tasks in real world applications, including service, agriculture, and healthcare where our society faces critical staff shortages. 

Visit Marc Biron's LinkedIn


Leading Edge Semiconductor Technology Drives Innovation and Market Growth

Tuesday, September 10



Semiconductor industry plays the essential role in driving the digital transformation. Recently, ChatGPT, Artificial intelligence, machine learning, … etc completely change the land scale of the innovations and solutions to improve the quality of life. Enormous computing performance with efficient energy consumption provided by leading edge technology made all those impossible in the past to become possible today. In this presentation, Paul de Bot will discuss the growth characteristics of the semiconductor market, which is expected to reach $1T by 2030, where the leading-edge technology (currently FinFET) is the largest growth driver across all market segments. The challenges of the leading-edge technology, including process and design solutions as well as design technology co-optimization (DTCO) will be discussed to tackle the megatrend in semiconductor industry.

Visit Paul De Bot's LinkedIn


Circuits and Systems for Embodied AI: Exploring uJ Multi-Modal Perception for Nano-UAVs on the Kraken Shield

Tuesday, September 10



Embodied AI requires pushing complex multi-modal models to the extreme edge for time-constrained tasks such as autonomous navigation of robots and vehicles. On small formfactor devices, e.g., nano-UAVs, such challenges are exacerbated by stringent constraints on energy efficiency and weight. In this paper, we explore embodied multi-modal AI-based perception for Nano-UAVs with the Kraken shield, a 7g multi-sensor (framebased and event-based imagers) board based on Kraken, a 22nm SoC featuring multiple acceleration engines for multi-modal event and frame-based inference based on spiking (SNN) and ternary (TNN) neural networks, respectively. Kraken can execute SNN realtime inference for depth estimation at 1.02k inf/s, 18 μJ/inf, TNN real-time inference for object classification at 10k inf/s, 6 μJ/inf, and real-time inference for obstacle avoidance at 221 frame/s, 750 μJ/inf. 

Visit Luca Benini's profile


Modeling of the MOSFET for the Design of Cryo-CMOS Circuits

Tuesday, September 10



This lecture highlights some of the challenges faced in the modeling of MOSFET devices operating at cryogenic temperature (CT). It will review the most important phenomena, including the saturation of the subthreshold swing (SS) below a critical temperature with its complex current composition, the increase of the threshold voltage, the self-heating effect, and finally the noise. Many circuits used for quantum computers and running at CT operate at RF. It is, therefore, important to understand how the MOSFET DC model can be extended to RF. Finally, we will show how the Gm/ID figure-of-merit (FoM) can help in designing Cryo-CMOS circuits when no compact models are available. All the presented models are backed up with experimental data acquired on devices from various advanced bulk, FDSOI, and FinFET CMOS technologies. 

Visit Christian Enz's LinkedIn


Computing with Physics – Theory, Devices and Circuit Design Perspective

Wednesday, September 11


BMCC (Auditorium A) 

This presentation introduces a physics-based computing paradigm and architecture, harnessing the collective dynamics of coupled oscillators to enable massive parallelism and energy-efficient computation. This approach overcomes the inherent limitations of classical von Neumann computing, enabling the execution of highly complex functions with remarkably low power consumption. At the core of this physics-based computing lies the interactive dynamics of devices within an open system, naturally minimizing their energy by transitioning to the ground state. The talk will cover computational theory pertaining to physical computing, as well as the materials and devices essential for its physical implementation. Physical computing leverages the intrinsic nonlinearities of devices and the memory of the physical system, enabling energy-efficient, in-memory computations with no data transfer, thus making it suitable as an Ising machines for solving NP-hard problems. Additionally, it serves as an energy-efficient hardware accelerator for AI applications. In conclusion, I will address the current challenges and advancements in energy-efficient computing.

Visit Aida Todri-Sanial's LinkedIn


Rethinking Mixed-Signal IC Design 

Thursday, September 12


BMCC (Auditorium A) 

Information and communication technologies require higher performance with lower power consumption. While new MOS technologies aim to extend Moore's law, scaling benefits have diminished. Heterogeneous integrated systems address some challenges, but the complexity of chiplets, including mixed-signal circuits, continues to rise. A shift in analog design is needed, emphasizing “full stack” solutions over traditional transistor-level focus to navigate trade-offs. This requires deeper system-level problem analysis and less focus on individual sub-blocks. 

Visit Gabriele Manganaro's LinkedIn

bottom of page