ADLINK Launches SMARC AI-on-Module to Drive Industrial AI at the Edge

ADLINK Launches SMARC AI-on-Module to Drive Industrial AI at the Edge

1101
Views
0 Comments

A future-proof, open standard formfactor of rugged industrial grade design with up to 15 years longevity and software portability through NXP® Semiconductors eIQAI SDK

ADLINK Technology Inc., a global leader in edge computing, has launched the LEC-IMX8MP SMARC module, the first SMARC rev. 2.1 AI-on-Module (AIoM) that uses NXP’s next-generation i.MX 8M Plus SoC for edge AI applications. The LEC-IMX8MP integrates NXP NPU, VPU, ISP and GPU computing in a compact size for future-proof AI-based applications across industrial AIoT/ IoT, smart homes, smart cities and beyond.

The powerful quad-core Arm® Cortex®-A53 processor runs up to 1.8 GHz with an integrated neural processing unit (NPU), delivering up to 2.3 Terra Operations Per Second (TOPS) for machine learning inference at the edge, suited for applications that require machine learning and vision systems paired with smart sensors to enable industrial decision-making.

“The i.MX 8M Plus AI-on-module in the SMARC form factor from ADLINK is great for industrial edge applications,” says Robert Thompson, Director of MPU Ecosystem, NXP. “We are trying to provide competitive solutions, such as this embedded module with AI capabilities, and the long-standing partnership and collaboration with ADLINK will enable us to continue driving more future market innovations.”

“As long-standing partners in an industry where our customers rely on us for top-of-the-line solutions, SMARC AI-on-Modules (AIOM) are the future of AI, with ADLINK at the forefront of this technology so critical to industrial edge applications,” says Henri Parmentier, Senior Product Manager, ADLINK. “Our LEC-IMX8MP SMARC 2.1 module will allow developers of AI-based smart embedded systems to implement cost-effective and future-proof designs built specifically for rugged environments that require higher-performance machine learning inference.”

The LEC-IMX8MP SMARC module features:

  • LVDS/DSI/HDMI graphic output, dual CAN bus/USB 2.0/USB 3.0, dual GbE ports (one with TSN), and I2S audio interface – in a low power envelope that is typically below 6W
  • Rugged design can sustain operating temperatures of -40°C to +85°C, and high shock and vibration environments for reliability in harsh industrial applications
  • Standard BSP support for Debian, Yocto and Android, including MRAA hardware abstraction layer (HAL), allows engineers to substitute modules, sensor HATs and port code written in Raspberry Pi or Arduino environments to the I-Pi
  • NXP eIQ machine learning software with consecutive inference on CPU cores, GPU cores and NPU. Support for Caffe, TensorFlow Lite, PyTorch and ONNX models. Enablement for models such as MobileNet SSD, DeepSpeech v1, and segmentation networks. Arm NN fully integrated into Yocto BSP, supporting i.MX 8

Delivering edge intelligence, machine learning and vision for a smart world, the LEC-IMX8MP SMARC 2.1 module is an excellent platform for AI-based applications, removing cloud dependency and preserving individual privacy. Targeted uses include smart homes and home automation, smart cities, logistics, healthcare diagnostics, smart buildings, smart retail, and industrial IoT including machine vision, robotics, and factory automation.

A ready-to-run I-Pi SMARC prototyping platform based on the LEC-IMX8MP module can be ordered online at ADLINK’s I-Pi SMARC theme and support site.

Find more information about the LEC-IMX8MP module here.

Watch the video.

Mike is the founder and editor of Electronics-Lab.com, an electronics engineering community/news and project sharing platform. He studied Electronics and Physics and enjoys everything that has moving electrons and fun. His interests lying on solar cells, microcontrollers and switchmode power supplies. Feel free to reach him for feedback, random tips or just to say hello :-)

view all posts by admin
Subscribe
Notify of
guest
0 Comments
Inline Feedbacks
View all comments
Archives