Ara240 product image.jpg   Welcome to the Ara240 DNPU Training! This page provides access to training materials, presentations, demos, recordings, and supporting resources related to the Ara240 DNPU. While live Q&A support will be available during the training period, all content will remain accessible for future reference and self-paced learning. Required Hardware FRDM i.MX 8M Plus Ara240 DNPU Ara240 16GB M.2 Module microSD >= 64GB storage USB-C debug cable Internet (Ethernet) HDMI Monitor 100 W power supply (Anker Charger is recommended) USB mouse USB keyword 1080p USB Camera Instructions Step1. Watch introduction DNPU Training Document Video Step2. Check the Hardware and Software Pre-requisites Getting Started Guide with FRDM-IMX8MPLUS Step3. Run the rest of the Software demo packages After completing the pre-work, each lab has its own guide document and a video guide you can use as support material in case you have any question at any step: Lab1: Running a GStreamer Pipeline for 8 Video Streams with YOLOv8n on Ara240 DNPU Ara Vision Multi-Stream YOLOv8 Object Detection Video Lab2: Running Unimodal Large Language Models on Ara240 DNPU LLM Edge Studio Video Lab3: Interacting with Vision-Language Models on Ara240 DNPU VLM Edge Studio Video Lab4: Enabling REST-Based LLM and VLM Inference on Ara240 DNPU eIQ AAF Connector Video

Watch Technical Overview Hands-On Training

This article provides you with general guidelines that apply for All Ara-2 Lab SW packages. Follow the steps below to download, prepare, and execute each lab successfully. 1. Download the Lab Package   All lab packages are hosted in the Ara2‑SW‑Packages v1.3.0 location. After opening the shared link: Navigate to corresponding Package folder Download all contents of that folder to your local machine Ensure the folder structure remains intact 2. Important Files Inside the Package   Each package includes two key HTML files: ️ INSTALL.html Provides step‑by‑step instructions to install prerequisites and prepare your environment. README.html Contains details about the lab scenario, execution steps, and additional notes. Important: These HTML files must be downloaded and opened locally. Opening them directly from the WorkDrive browser will cause them to load incorrectly. Once downloaded, simply double‑click to open them in your browser — they are fully self‑contained and easy to follow. 3. Running the Lab   After completing the installation steps in INSTALL.html: Review the scenario and steps in README.html Execute the lab tasks at your own pace Return to the forum to share results, ask questions, or support others! 4. Watch the Walkthrough Video   A video file is included in each package. This video walks you through the same steps described in INSTALL.html and README.html, providing a visual guide to installation, setup, and execution. You can use the video as: A companion guide while following the HTML instructions A quick overview before starting the lab A reference if you prefer visual, step‑by‑step walkthroughs    

Applies to: LLM Edge Studio Go Point Demo Workaround: Requires custom llm_params configuration The model Qwen2.5-Coder-1.5B is not supported out-of-the-box in LLM Edge Studio using the default configuration. The model loading will fail when attempting to run the model with the default  server_config.json fix: update the  llm_params  section in  server_config.json  with compatible sampling values: - Open config file: vi llm-edge-studio/usr/share/llm-edge-studio/server_config.json - Replace the parameters as follows: "llm_params": { - "temperature": 0.0, - "top_k": 0, - "top_p": 0.0, + "temperature": 1.0, + "top_k": 50, + "top_p": 0.95 } Qwen2.5-Coder-1.5B should run with proper output quality and sampling behavior after applying the updated configuration.  Remarks:  This configuration enables proper behavior for Qwen2.5-Coder-1.5B It may generate warning messages when running Qwen2.5-Instruct-7B, but the model will continue to function normally BR,

Applies to: LLM / VLM Edge Studio Go Point Demo Workaround: Update demos via deb package In the current BSP release, users may encounter a runtime failure when the requesting inference metrics after running an inference in the LLM / VLM Edge Studio demo, as a result, metrics are not displayed at all. This issue is caused by a change in the metrics response format from the AAF Connector, which is not compatible with the current demo version.  To fix, update the demo by installed the patched Debian package attached to this thread.  - Download and copy deb packages to target board:  scp vlm-edge-studio_1.0.1.deb [email protected]: scp llm-edge-studio_2.0.1.deb [email protected]: - Overwrite the demo installation: dpkg -i --force-overwrite vlm-edge-studio_1.0.1.deb dpkg -i --force-overwrite llm-edge-studio_2.0.1.deb BR,

  Applies to: VLM Edge Studio Demo using Qwen‑VL backend Workaround: Apply the attached  apply_qwenvl_fix.sh  script   In the current VLM Edge Studio Demo, users may encounter a runtime failure when starting the Qwen‑VL Vision Language Model. The issue is caused by incorrect processor initialization, where the code attempts to load tokenizer and image processor assets directly from a hard‑coded Hugging Face model reference instead of the locally deployed tokenizer path. This typically manifests as: Model startup failures Tokenizer / processor load errors Inconsistent behavior when running in offline environments To fix no configuration changes are required, copy and apply the provided patch on the board: root@imx95evk:~# chmod a+x apply_qwenvl_fix.sh root@imx95evk:~# sh apply_qwenvl_fix.sh remarks: please make sure the path in the script match your installation: PROD_FILE="/usr/share/eiq/aaf-connector/venv/lib/python3.13/site-packages/eiq_aaf_connector/llm_engines/QwenVL.py"

Introduction DNPU Training   Explore the evolution of AI at the edge, from perception AI and generative AI to the emerging era of Agentic AI. This session introduces NXP’s discrete NPU strategy and examines how neural networks, transformer models, and AI agents enable intelligent decision-making for industrial, healthcare, and automation applications. Participants will gain an understanding of the technologies driving the next generation of intelligent edge systems.