Skip to main content

NXP i.MX 8MP EVK

i.MX 8MP EVK

The i.MX 8M Plus Evaluation Kit is a powerful and flexible platform for edge AI development. With a quad-core Cortex-A53 processor and an integrated neural processing unit (NPU), it enables real-time ML inference without overloading the CPU—ideal for vision, voice, and industrial applications.

Combined with Avocado OS, teams can deploy complex workloads quickly using a consistent, production-ready Linux environment.

Technical Specifications

ComponentDetails
CPUQuad-core Arm Cortex-A53 (1.8 GHz)
MCUSingle Arm Cortex-M7 (800 MHz)
NPUi.MX 8M Plus SoC integrated NPU
AI PerformanceUp to 2.3 TOPS (INT8)
Memory6GB 32-bit LPDDR4
Memory Bandwidth12.8 GB/s
Storage32GB eMMC
ConnectivitySingle M.2 Key E wireless module with Wi-Fi and Bluetooth
Power Modes1.7W / 2W / 4W / 6W

Getting Started

Get up and running with the Avocado Linux SDK for i.MX 8M Plus in minutes.

Prerequisites

  • A Mac (macOS 10.12+) or Linux (Ubuntu 22.04+, Fedora 39+) development machine
  • Docker installed
  • 10GB+ available disk space

Provisioning your device

  1. Initialize your project
avocado init --target imx8mp-evk avocado-imx8mp
cd avocado-imx8mp
  1. Install all components (SDK, extensions, and runtime dependencies):
avocado install -f
  1. Build all components
avocado build
  1. Provision a runtime
avocado provision -r dev
  1. Verify that a complete system image file was output:
ls -l /opt/_avocado/output/avocado-image-imx8mp*.wic

Flashing the Image

  1. Insert a microSD card (minimum 8GB) into your development machine

  2. Identify the device (typically /dev/sdX or /dev/mmcblkX):

lsblk
  1. Flash the image to the SD card:
sudo dd if=/opt/_avocado/output/avocado-image-imx8mp*.wic of=/dev/sdX bs=4M conv=fsync status=progress
  1. Insert the SD card into the i.MX 8MP EVK and power on

Booting and Connecting

  1. Connect to the serial console (115200 8N1):
minicom -D /dev/ttyUSB0 -b 115200
  1. Default login credentials:

    • Username: root
    • Password: (no password by default)
  2. Configure networking and start services as needed

NPU Acceleration

The i.MX 8M Plus includes a dedicated Neural Processing Unit (NPU) providing 2.3 TOPS of AI acceleration. This enables running complex ML models directly on the edge device.

Supported Frameworks

  • TensorFlow Lite
  • ONNX Runtime
  • PyTorch (via ONNX)
  • OpenVINO

Example: Running Object Detection

# Install ML runtime packages
avocado-repo sysext install tensorflow-lite-vx-delegate -y

# Build and deploy with NPU support
avocado-build sysext tensorflow-lite-vx-delegate

# On target, run inference benchmark
cd /usr/share/tensorflow-lite/examples
./benchmark_model --graph=mobilenet_v2.tflite --use_nnapi=true

Real-Time Processing with Cortex-M7

The i.MX 8M Plus includes a dedicated Cortex-M7 core running at 800 MHz for real-time processing tasks.

Loading Firmware to M7 Core

# Stop the M7 core
echo stop > /sys/class/remoteproc/remoteproc0/state

# Load firmware
echo hello_world.elf > /sys/class/remoteproc/remoteproc0/firmware

# Start the M7 core
echo start > /sys/class/remoteproc/remoteproc0/state

Dual Camera Support

The EVK supports dual MIPI CSI cameras for stereo vision and multi-angle capture applications.

Camera Pipeline Setup

# List available cameras
v4l2-ctl --list-devices

# Capture from camera 0
gst-launch-1.0 v4l2src device=/dev/video0 ! video/x-raw,width=1920,height=1080 ! autovideosink

# Simultaneous dual camera capture
gst-launch-1.0 \
v4l2src device=/dev/video0 ! queue ! autovideosink \
v4l2src device=/dev/video1 ! queue ! autovideosink

Industrial Networking

The i.MX 8M Plus EVK includes Time-Sensitive Networking (TSN) support for deterministic industrial communication.

TSN Configuration

# Configure TSN on eth1
ip link set dev eth1 up
tc qdisc add dev eth1 parent root handle 100 taprio \
num_tc 3 \
map 2 2 1 0 2 2 2 2 2 2 2 2 2 2 2 2 \
queues 1@0 1@1 2@2 \
base-time 1000000000 \
sched-entry S 01 300000 \
sched-entry S 02 300000 \
sched-entry S 04 400000 \
clockid CLOCK_TAI

Power Management

The i.MX 8M Plus supports multiple power modes for optimizing energy consumption.

Power Mode Configuration

# Check current power mode
cat /sys/power/state

# Enter suspend mode
echo mem > /sys/power/state

# Configure CPU frequency scaling
cpufreq-set -g powersave

Security Features

The i.MX 8M Plus EVK includes comprehensive security features:

  • High Assurance Boot (HAB)
  • Cryptographic acceleration (CAAM)
  • Secure boot chain
  • TrustZone support
  • Inline encryption engine

Enabling Secure Boot

Secure boot configuration is handled through the Avocado build system:

# Build with secure boot enabled
AVOCADO_SECURE_BOOT=1 avocado-build image

Additional Resources