AI Meets Functional Safety: Exploring Ara DNPU Potential

AI Meets Functional Safety: Exploring Ara DNPU Potential

AI Acceleration Week (Ep. 2): Functional Safety + Edge AI with Ara 240 and F&S Elektroniksystem In the second episode of EdgeVerse Techcast's AI Acceleration Week, co-hosts Bridgette Stone and Kyle speak with Andreas "Andy" Kopitz of F&S Elektroniksystem about how developers can start building edge AI products using NXP's Ara 240 alongside the i.MX 95, especially for robotics and autonomous systems. Andy explains that customers increasingly want functional safety, Linux, ROS 2, multiple high-resolution camera streams, and real-time AI decision-making on compact platforms, creating a gap in AI performance that Ara 240 helps close. He describes how pairing i.MX 95 for system control, security, connectivity, HMI, and safety architecture with Ara 240 for AI acceleration enables demanding applications, and how F&S supports developers with complete platforms, an Ara 240 M.2 module, Linux/Yocto BSPs, and functional-safety approaches using hypervisor and safety supervisor layers with direct engineering support. Episode Resources: * F&S M.2 Edge AI Ara 240 SMARC (Video) [https://youtu.be/aXShuAWl154?si=Eo0rQA6n9py_y713] * M2 Ara2 Edge AI Product Page [https://www.fs-net.de/en/embedded-modules/m2-ara2-edgeai/] * www.nxp.com/Ara240 [https://www.nxp.com/Ara240] 00:00 Welcome to AI Week 01:05 Meet Andy from F&S 01:59 Hobbies and Banter 03:21 Austin Meeting Insights 04:11 Customer Demands and Gaps 06:35 Ara 240 Closes the Gap 08:14 Why F&S Was Chosen 09:38 F&S i.MX95 plus Ara Platform 11:12 Software and Linux Safety 13:11 Robotics with ROS2 and Safety 14:45 Getting Started Resources 16:17 Wrap Up and Next Episode

Kickstart AI Acceleration Week: Unveiling DNPU's Role in AI Applications

AI Acceleration Week Kickoff: Edge AI Trends, Multimodal & Generative AI, and NXP DNPUs with Ravi Annavajjala Edge Verse Techcast kicks off AI Acceleration Week with the first of three episodes exploring technologies shaping the future of edge AI. Hosts Bridgette and Kyle welcome Ravi Annavajjala, former CEO of Kinara and now leader of NXP's NPU product line. Ravi explains key edge AI trends, including the shift from cloud-centric AI to distributed intelligence driven by latency, privacy, bandwidth, and real-time responsiveness. The episode also covers the evolution from vision-only AI to multimodal systems combining voice, video, sensors, and context. Ravi discusses why edge inferencing matters for robotics, industrial automation, and automotive, and introduces NXP's discrete neural processing units (DNPUs) as dedicated accelerators for advanced inference workloads like generative and multimodal AI. The episode also highlights NXP's eIQ software environment and ecosystem partners enabling developers to move from prototyping to production. Episode Resources: * www.nxp.com/ara240 [https://www.nxp.com/ara240] * www.nxp.com/eiq [https://www.nxp.com/eiq] * www.nxp.com/ai [https://www.nxp.com/ai] 00:00 AI Acceleration Week Intro 01:35 Meet Ravi Annavajjala 02:50 Big Trends in Edge AI 04:18 Why Edge Beats Cloud 05:27 Kinara Joins NXP 06:50 What Is a DNPU 08:10 Top Edge AI Applications 08:40 Tools and Ecosystem Support 10:30 Bonus AI Tennis Referee 11:46 Key Takeaways and Next Episodes

Accelerate Your Development: The Role of Distribution FAEs

Why NXP Distribution FAEs Matter: Faster Designs, Better Support Kyle Dando and Bridgette Stone explain the role of NXP's distribution field application engineers (DFAEs) and why NXP invests in annual distributor FAE training, recently held in Austin. They describe how distributors like Avnet, Arrow, Digi-Key, Future, and Mouser provide sales, inventory and pricing management, customer credit, and critical technical support, with DFAEs often serving as the first technical contact and a bridge to NXP subject matter experts. The episode recaps the four-day Americas training attended by over 100 engineers, featuring lectures and breakout sessions across pillars such as motor control, AI/ML, Security, wireless, and automotive. Plus hands-on training with Visual Studio Code, Configuration Tools, SEC Provisioning, and Zephyr development. They emphasize the "train-the-trainer" model and share an example of an FAE quickly connecting a new PMIC to a customer's secure lock project. Episode Resources: * Avnet Engineering Support [https://www.avnet.com/apac/solutions/design/engineering-support/electronic-components-engineering-support/] * Arrow Engineering Support [https://www.arrow.com/en/services/engineering-and-design.html] * Future Electronics Engineering Support [https://www.futureelectronics.com/about-future/advanced-engineering] * Digi-Key Engineering Support [https://forum.digikey.com/] * Mouser Engineering Support [https://www.mouser.com/technical-resources/] 00:00 Welcome to EdgeVerse 00:31 What DFAEs Do 02:01 Distributor Roles Explained 03:44 Inside FAE Training 04:43 Why Use Distribution 07:20 Cool Tech Highlights 09:02 How Training Helps You 09:37 Real World Success Story 11:11 Workshops and Next Steps 11:29 Wrap Up and Call to Action

From Data to Deployment: Mastering Edge AI with eIQ

NXP eIQ Explained: From Edge AI Deployment to Agentic AI Autonomy Bridgette Stone and Kyle Dando host Ted Kao to break down NXP's eIQ (Edge Intelligence) AI/ML development environment for NXP MCUs and application processors, aimed at simplifying the path from models to efficient on-device deployment. Ted outlines key components like the eIQ AI Toolkit, support for inference runtimes including TensorFlow Lite, ONNX Runtime, Arm, and ExecuTorch, plus integrations with MCUXpresso and Yocto. He explains two entry paths—bring your own model (eIQ Toolkit/Neutron SDK/AI Hub) and bring your own data (Time Series Studio with AutoML for time-series use cases)—and mentions hardware testing via FRDM boards and a planned remote board farm. The episode highlights newer additions such as the eIQ Agentic AI Framework, eIQ AI Hub, and Gen AI Flow, and points listeners to nxp.com/ai and nxp.com/eIQ to get started. Episode Resources: * www.nxp.com/ai * www.nxp.com/eiq 00:00 Welcome to EdgeVerse 00:28 Why Edge AI Matters 01:45 Meet Ted Kao 02:31 What eIQ Is 04:24 Inside the eIQ Ecosystem 05:37 Two Paths to Start 07:01 Bring Your Own Data 08:24 Testing on Real Hardware 09:00 New Tools and Agentic AI 11:28 Getting Started Resources 12:33 Wrap Up and Subscribe

NXP's Block Diagram Designer: Kickstart Your Next Project Today

Kickstart Your Next Design Faster with NXP's Block Diagram Designer | EdgeVerse Techcast In this EdgeVerse Techcast episode, co-hosts Kyle Dando and Bridgette Stone welcome Rebecca Phillips from NXP's digital online experience team to introduce NXP's new Block Diagram Designer, a web-based tool that helps engineers move from concept to design faster by turning the "blank page" into a structured, component-based block diagram. Rebecca explains how engineers can start from reusable application templates or a blank canvas, get assistance choosing compatible components using built-in design rules, and export designs to industry-standard EDA tools like Eagle, Altium, and Autodesk to speed handoff into implementation. The tool is powered by technical documentation and design files organized as "CUBOS," covering NXP and required non-NXP parts, with datasheets and related resources linked in one place. Episode Resources: * www.nxp.com/block-diagram-designer [https://www.nxp.com/block-diagram-designer] * Create a new Diagram [https://nxp.celus.io/design-studio] 00:00 Welcome to EdgeVerse 01:10 Meet Rebecca Phillips 01:54 FIRST Robotics Tangent 03:01 What Is Block Diagram Designer 04:35 NXP Customer Journey 06:20 Top Three Features 08:29 Inside the CUBOS Data 11:39 Future AI Assistant 12:34 How to Get Started 13:29 Wrap Up and Subscribe