Sony Semicon (EU)’s Post

Technology Report: The State of Edge AI in 2024 (Free Download) I am delighted to have contributed to and co-authored the latest Wevolver (Jess Miley Bram Geenen Tasos Polygenis) in collaboration with Samir Jaber (editor in chief of the report) and Ravi Rao. The report combines months of research into Edge AI applications across industries, presenting industry-specific insights, real-life case studies, and a peek into generative AI technology and its convergence with edge computing. The report is developed in partnership tinyML Foundation and sponsored by several industry leaders in #edgeAI including Renesas Electronics, Synaptics Incorporated, Arduino, Nordic Semiconductor, Syntiant Corp., Mouser Electronics, and others. You can download your copy of the report for free here: https://lnkd.in/ekJPm7ST Rute C. Sofia Oscar Lazaro George Kousiouris Dimosthenis Kyriazis George Makridis Damir Filipovic #EdgeAI #AI #TechnologyReport #EdgeAIReport #2024EdgeAiReport #generativeAI

Neuromorphic Analog Signal Processing (NASP) is pushing the boundaries of ML technology. This technology mimics the human nervous system and accelerates computation while optimizing power use, paving the way for unprecedented efficiency. The application-specific AI models, powered by NASP, are designed for the analog circuitry to resemble neurons and axons, eliminating the need for an analog-to-digital converter. This innovation allows for direct AI computations next to sensors, reducing latency associated with cloud-based processing. Enhancements to Edge AI, such as NASP and embedding techniques, have proven potent solutions for Industrial Internet of Things (IIoT) applications. Processing sensor signals directly on the chip shrinks raw data by an incredible more than 1,000x, enabling real-time insights. This technology is not just addressing power and latency issues but truly reshaping the landscape of edge computing.  https://lnkd.in/gR5Seiue #NASP #TinyML #EdgeComputing #InnovationUnleashed

Technology Report: The State of Edge AI in 2024 (Free Download) I am delighted to have contributed to and co-authored the latest Wevolver (Jess Miley Bram Geenen Tasos Polygenis) in collaboration with Samir Jaber (editor in chief of the report) and Ravi Rao. The report combines months of research into Edge AI applications across industries, presenting industry-specific insights, real-life case studies, and a peek into generative AI technology and its convergence with edge computing. The report is developed in partnership tinyML Foundation and sponsored by several industry leaders in #edgeAI including Renesas Electronics, Synaptics Incorporated, Arduino, Nordic Semiconductor, Syntiant Corp., Mouser Electronics, and others. You can download your copy of the report for free here: https://lnkd.in/ekJPm7ST Rute C. Sofia Oscar Lazaro George Kousiouris Dimosthenis Kyriazis George Makridis Damir Filipovic #EdgeAI #AI #TechnologyReport #EdgeAIReport #2024EdgeAiReport #generativeAI

AI Frontiers - I recommend attending this Webinar next Thursday 24 October 2024, 2pm- 3pm. Topics include energy-efficient AI hardware solutions and novel semiconductor technologies and neuromorphic computing architectures.

#EW24 has concluded, and we're excited to share a brief 4-minute video featuring Julia Munoz, our Director of Business Development, and Lisa Trollo, ST MEMS Sensors Ecosystem and Digital Marketing Manager at STMicroelectronics. In this video, they explore the key pillars of in-sensor AI architecture, discussing its advantages for resource-constrained and always-on devices through the use of #STMicroelectronics IMUs with intelligent sensor processing units and ultra-compact Neuton.AI neural networks. #STAuthorizedPartner #STPartnerProgram Danilo Pietro Pau, Marco Bianco, Remi El-Ouazzane, Francesco Doddo, Philip Lolies, Mahesh Chowdhary, Matteo Fusi, Ernesto Manuel Cantone, Edoardo Gallizio, Sarah Zerbini, Sarah Bracken, Miguel CASTRO, Marco Angelici, Michele Ferraina #Embeddedworld #tinyml #aibasedsolutions #ondeviceai  #neuralnetworks #ml #insensorai #smartsensors #remotecontrol #gesturerecognition #aifeatures #MEMS #Sensors #Future #Innovation #Technology #smartring #activityrecognition #trackingsolution #IMU #ISPU

Lightmatter: Revolutionizing Computing with Light-Speed Tech Are you ready for the future of computing? Lightmatter, a pioneering MIT spin-off, is using photonics to break through limitations and achieve lighting-fast data processing ⚡️ Their secret weapon? Combining the power of light (photons) with electricity (electrons) for groundbreaking efficiency. This results in: > Energy-saving AI: Lightmatter's Envise chip tackles complex AI tasks with significantly lower power consumption > Blazing-fast data transfer: Passage, their interconnect chip, uses light for near-instantaneous communication between processors, similar to fiber optics ⏩ > Sustainable future: This innovative approach combats the ever-growing energy demands of data centers and AI Lightmatter is poised to disrupt the computing landscape, paving the way for a greener and faster future. Are you ready to join the journey? https://lnkd.in/gYHjHYtr #Lightmatter #Photonics #AI #Sustainability #MIT #FutureofComputing #PhtonicSpots #PhotonicsTimes #optics Massachusetts Institute of Technology Image Caption: Lightmatter's Passage chip, inspired by fiber optics, uses light to supercharge communication between processors for faster AI and cloud computing. (Credits: Courtesy of the researchers. Edited by MIT New)

📢 Join Marcellino Gemelli at the MEMS and Sensors Executive Conference (MSEC) 2024 in Quebec, Canada, on October 8, at 9:00 a.m. for his talk on "The Seven Steps to Bring AI Tools Inside a Semiconductor Company." 📅 MSEC 2024 gathers leaders from the #MEMS and #sensor supply chain, focusing on business updates, strategic trends, and key technologies. As pioneers in adopting new technologies, especially #AI, we've spent over a decade using embedded algorithms for edge AI to enhance performance and privacy of our sensor data. Recently, new AI tools are reshaping our operations, from shortening software design cycles to improving engineering productivity and customer interactions. Marcellino will share insights on this transformation in seven steps. Curious? Learn more here 👉 https://bit.ly/3N6YQAK.

#news - We are researching ways to make sensors more high performance, more energy-efficient and smarter. ⚡📈 Through approaches such as edge computing, in which data processing takes place directly at the sensor, we are also helping to increase the trustworthiness of systems. 🖥️🔒 In a number of projects, we have succeeded in opening up new applications by combining sensor technology with artificial intelligence. 🤖📊 Find an overview of our projects in our annual report: 👉 https://lnkd.in/dTBT9vC --- #Fraunhofer #FraunhoferIPMS #whatsnext #SensorTechnology #EdgeComputing #EnergyEfficiency #HighPerformance #SmartSensors #ArtificialIntelligence #AI #Innovation #TechResearch #TrustedElectronics #Sensors

🌟 Pioneering the Future: TinyAI and Its Groundbreaking Potential 🚀 Pioneering efforts have led to the development of AI compression and acceleration techniques like knowledge distillation, fusion, pruning, and quantization, alongside the creation of compact AI algorithms that use minimal trainable parameters while maintaining strong inference performance. Simultaneously, specialized AI hardware, including neuromorphic chips, FPGAs, NPUs, ultra-low-power DNN accelerators, and optoelectronic neural networks (ONNs), has been designed to enhance AI capabilities in resource-constrained devices. 💡 This perspective delves into the core principles of TinyAI and its supporting hardware, mapping out future research and innovation pathways. It also explores the vast potential applications of TinyAI, extending beyond current uses to forecast transformative advancements in edge computing, setting the stage for a new era of AI-powered technology.

Counting down to Electronica in Munich next week! 🎉 I can't wait to connect with industry leaders, discover the latest in electronics, and see emerging trends up close. Let's connect if you're looking to deploy AI models optimized for resource-constrained devices to ensure accuracy, efficiency, and scalability! #Electronica2024 #Semicon #Munich #TechInnovation #Networking #Embedded #EdgeAI #AI