Louis Besème

Meet Sylvie Menezo, the founder and CTO of SCINTIL Photonics, a deep-tech semiconductor startup spun out of CEA-Leti in Grenoble. Her path is a story of scientific ambition, technical excellence, and bold entrepreneurship: building a platform that brings together lasers, modulators, detectors and waveguides into a single manufacturable chip using their proprietary SHIP (Scintil Heterogeneous Integrated Photonics) process. Scintil’s flagship product - LEAF Light, a DWDM-native single-chip light engine - delivers 6.4 Tbps/mm edge bandwidth with much lower power consumption than traditional optical solutions 🚀 . Data centres are hitting bandwidth, latency, and energy efficiency constraints - and the shift from copper to integrated photonics is a foundational infrastructure transformation. Scintil helps Europe compete in a critical technology stack, reduce reliance on overseas optical component supply chains, and drive the next generation of compute and communications systems. Scintil closed a $58 million (≈ €50 million) Series B round, led by Yotta Capital Partners and NGP Capital, with strategic participation from NVIDIA, BNP Paribas Développement, and others including existing backers like Supernova Invest, Bpifrance Digital Venture, Bosch Ventures, INNOVACOM, and Applied Ventures. 🇪🇺 European Innovation Council and SMEs Executive Agency (EISMEA) has also supported Scintil earlier, helping with early R&D, validation, and bridging capital. 🇪🇺 Europe is not just a place for innovation - it is the place to grow deep tech infrastructure and sovereign capacity. Europe is ready to support dozens more founders like Sylvie - especially women in deep tech - with the capital, networks, and public backing they need 🚀

181

3 commentaires

SYMPLECTIC AUTOENCODER LEARNING HAMILTONIAN DYNAMICS AT SCALE: A DIFFERENTIAL-GEOMETRIC APPROACH by Ecole normale supérieure ABSTRACT By embedding physical intuition, network architectures enforce fundamental properties, such as energy conservation laws, leading to plausible predictions. Yet, scaling these models to intrinsically high-dimensional systems remains a significant challenge. This paper introduces Geometric Reduced-order Hamiltonian Neural Network (RO-HNN), a novel physics-inspired neural network that combines the conservation laws of Hamiltonian mechanics with the scalability of model order reduction. RO-HNN is built on two core components: a novel geometrically-constrained symplectic autoencoder that learns a low-dimensional, structure-preserving symplectic submanifold, and a geometric Hamiltonian neural network that models the dynamics on the submanifold. Our experiments demonstrate that RO-HNN provides physically-consistent, stable, and generalizable predictions of complex high-dimensional dynamics, thereby effectively extending the scope of Hamiltonian neural networks to high-dimensional physical systems.

129

5 commentaires

The boundary between fundamental research and industrial applications in quantum science has never been thinner. In today’s Le Monde opinion piece, Christophe Jurczak (Quantonation) joins Antoine Browaeys (CNRS), Loïc Henriet (Pasqal), and Nicolas Treps (Sorbonne Université) to advocate for stronger collaboration between academic labs and start-ups. Read here : https://lnkd.in/eAZ2w9_A Their message is clear: it is within this continuous dialogue between science and entrepreneurship that the most fertile innovation emerges – and where a solid quantum industry can be built, serving society and advancing technological leadership.

98

6 commentaires

Why engage youth in science? 🧑‍🎓 On 25 September, CERN hosted an exclusive preview screening at #CERNScienceGateway of the documentary “Les vrais chercheurs ne savent pas ce qu’ils cherchent…” directed by Jean-Luc Perreard and Cédric Klapisch. The film explores the mysteries of the #HiggsBoson and the #BigBang, celebrating research as a journey of experimentation, emotion, and the transfer of knowledge across generations. Hear from the directors and special guests – #CERN Director-General Fabiola Gianotti, architect Renzo Piano, and John Elkann, President of Fondazione Agnelli – on why it is important to get the young generation interested in science.

295

14 commentaires

Good news for high-energy physics — thanks to our renewed SCOAP³ agreement with CERN, authors can keep publishing open access in Physical Review Letters, Physical Review C, and Physical Review D — with no fees and no interruption — through 2027. Learn more: https://go.aps.org/352Nz1X #OpenScienceWeek

112

2 commentaires

Extremely cool work by researchers at Ecole normale supérieure and Meta on understanding the representational similarity between human brains and AI models! More specifically, they trained DINOv3 self-supervised vision transformers of different sizes, with different training amounts, and different image types, and compare their representations to that of the human brain recorded with fMRI MEG. They found that size matters (largest models with the most human-centric images had the highest brain similarity scores), and during training the models first aligned with early representations of sensory cortices followed by alignment with late and prefrontal representations with a lot more training data. This is fascinating in not only better understanding deep neural networks but also can be a powerful tool for understanding the human brain. #meta #ecole #normale #superieure #research #understanding #brain #neuralnetwork #ai #model #similarity #ai #ml #dl #artificialintelligence #machinelearning #deeplearning #transformers #human #fmri #meg #cool #fascinating

444

11 commentaires

Nonlinear control, electric vehicles, complex dynamical systems, fault diagnosis, and intelligent energy systems — these are among the most pressing topics at the heart of today’s energy transition and sustainable mobility efforts. I'm pleased to share a collection of recently published volumes, the result of international scientific collaboration. These works address key challenges with both rigorous theoretical foundations and practical application insights: -Autonomous Electric Vehicles: Nonlinear Control, Traction, and Propulsion -Intelligent Control for Electric Power Systems and Electric Vehicles -Fault Diagnosis for Electric Power Systems and Electric Vehicles - Nonlinear Optimal and Flatness-based Control Methods and Applications for Complex Dynamical Systems - Control and Estimation of Dynamical Nonlinear and Partial Differential Equation Systems: Theory and Applications These books offer advanced theoretical tools and computational techniques applied to: Autonomous electric vehicles and dynamic propulsion control Intelligent fault diagnosis in power and transportation systems Nonlinear optimal control and modeling strategies Smart integration of automation, energy systems, and control theory This collection is a valuable resource for researchers, PhD students, and professionals working at the intersection of automation, smart energy, and intelligent mobility. A sincere thank you to Dr. Gerasimos Rigatos to have made this possible and for his commitment to advancing research in areas vital to sustainability and innovation. #ElectricVehicles #SmartGrids #NonlinearControl #EnergyTransition #IntelligentSystems #AutonomousVehicles #ControlTheory #AcademicResearch #SustainableMobility

226

17 commentaires

Our new state-of-the-art AI model Aeneas transforms how historians connect the past. 📜 Ancient inscriptions often lack context – it's like solving a puzzle with 90% of the pieces lost to time. Aeneas helps researchers interpret and situate inscriptions in their past context. By transforming each ancient text into a unique historical fingerprint, Aeneas can identify similarities across 176,000 Latin inscriptions. In our study, historians found these ‘parallels’ to be helpful research starting points 9 out of 10 times - improving their confidence by 44%. We tested Aeneas on the Res Gestae Divi Augusti – one of the most debated ancient inscriptions. Without prior knowledge, it successfully mapped out the leading scholarly theories on its dating, showing how AI can help model history in a quantitative way. 📊 Aeneas is now accessible through: 👉A website for researchers 🧑‍💻Open-source code and dataset 📚Syllabus for classrooms 🏛️Upgraded Ithaca ancient Greek model We’re excited to see how more people use this work to uncover the past. Find out more → https://goo.gle/4kVkh6n

1 452

85 commentaires