🚀 Introducing COAC: Cross-Layer Optimization for Efficient CNN Processing on Edge Devices ⚡📱 Authors: Steven Colleman, Man Shi, Marian Verhelst Published in: IEEE Transactions on VLSI Systems, July 2023 CNNs are becoming increasingly complex, making efficient deployment on

🚀 Analog vs. Digital In-Memory Computing: Which Leads the Edge? 🤔⚡ In-Memory Computing (IMC) is revolutionising machine learning at the edge with its promise of energy, throughput, and area efficiency. But how do Analog IMC (AIMC) and Digital IMC (DIMC) really stack up? 📊🔍

📢 Stay Updated with Us! 📰 🌟 Subscribe to our newsletter and be the first to discover our latest papers, research breakthroughs, and exciting activities! 🎉 📚 🧠 Let's keep the knowledge flowing. 👉 convolve.eu #Newsletter #Research #Innovation #StayInformed

🚀 The Future of European Semiconductors Starts Now! 🇪🇺🔬 With new US tariffs on microchips disrupting global supply chains, Europe’s need for semiconductor independence has never been clearer. The CONVOLVE Project is leading the way, pioneering adaptive computing to create

🎯 Research: Tailored Mixed-Precision Quantization for GRUs Excited to share our work: "Towards a tailored mixed-precision sub-8-bit quantization scheme for Gated Recurrent Units using Genetic Algorithms" by Riccardo Miccini, Alessandro Cerioli, Clément Laroche, Tobias

📢 11 New Publications from Project Convolve! 🚀 We're excited to share 11 new research publications from the Convolve project, covering a range of cutting-edge topics in AI, hardware acceleration, secure compilation, and more. Check them out! 📄 Featured Publications: 🔹

🔧 Bridging compiler prototyping and production — toward the flexible, modular toolchains envisioned in CONVOLVE. This new paper introduces xDSL, a lightweight compiler framework designed to work alongside MLIR, enabling rapid prototyping, teaching, and research in compiler

🔗 Toward composable, scalable compilers for the future of computing. This new paper explores how to build modular, reusable compiler components by composing dialects in multi-level IRs. The authors propose mechanisms for defining and combining dialects in a principled way, with

🚀 At this year’s OSSMPIC, Prof. Henk Corporaal delivered a compelling invited talk: "CGRAs for the Edge: Balancing Compute Efficiency and Flexibility." His presentation showcased the SOTA in CGRA development, including key outcomes from the CONVOLVE project, like the

🔐 CONVOLVE at DATE 2025 – Lyon We were excited to showcase our work on secure edge-AI architectures at the @DateConference as part of the Multi-Partner Project track. Our team presented a 15-minute talk and hosted an interactive poster session on the security architecture of

💡 How can Europe lead in ultra-low-power secure processors for edge computing? At the NEUROCOM workshop during HiPEAC 2025, a high-level panel discussion brought together experts from academia, industry, and policy to explore Europe’s position in this fast-evolving field. 🎙️

🌍 CONVOLVE at ETCEI 2025 We are pleased to share that CONVOLVE was represented at the Emerging Tech Conference on Edge Intelligence (#ETCEI2025), held in Athens 🇬🇷. This year’s conference, organised by Hetia - Hellenic Emerging Technologies Industry Association, University of